Glide User Manual

Glide User Manual
Glide User Manual
Glide 6.7
User Manual
Schrödinger Press
Glide User Manual Copyright © 2015 Schrödinger, LLC. All rights reserved.
While care has been taken in the preparation of this publication, Schrödinger
assumes no responsibility for errors or omissions, or for damages resulting from
the use of the information contained herein.
Canvas, CombiGlide, ConfGen, Epik, Glide, Impact, Jaguar, Liaison, LigPrep,
Maestro, Phase, Prime, PrimeX, QikProp, QikFit, QikSim, QSite, SiteMap, Strike, and
WaterMap are trademarks of Schrödinger, LLC. Schrödinger, BioLuminate, and
MacroModel are registered trademarks of Schrödinger, LLC. MCPRO is a trademark
of William L. Jorgensen. DESMOND is a trademark of D. E. Shaw Research, LLC.
Desmond is used with the permission of D. E. Shaw Research. All rights reserved.
This publication may contain the trademarks of other companies.
Schrödinger software includes software and libraries provided by third parties. For
details of the copyrights, and terms and conditions associated with such included
third party software, use your browser to open third_party_legal.html, which is in
the docs folder of your Schrödinger software installation.
This publication may refer to other third party software not included in or with
Schrödinger software ("such other third party software"), and provide links to third
party Web sites ("linked sites"). References to such other third party software or
linked sites do not constitute an endorsement by Schrödinger, LLC or its affiliates.
Use of such other third party software and linked sites may be subject to third
party license agreements and fees. Schrödinger, LLC and its affiliates have no
responsibility or liability, directly or indirectly, for such other third party software
and linked sites, or for damage resulting from the use thereof. Any warranties that
we make regarding Schrödinger products and services do not apply to such other
third party software or linked sites, or to the interaction between, or
interoperability of, Schrödinger products and services and such other third party
software.
May 2015
Contents
Document Conventions .................................................................................................... vii
Chapter 1: Introduction ....................................................................................................... 1
1.1 Running Schrödinger Software .............................................................................. 1
1.2 Starting Jobs from the Maestro Interface ............................................................. 3
1.3 Using Glide ................................................................................................................. 4
1.4 Documentation........................................................................................................... 5
1.5 Citing Glide in Publications ..................................................................................... 5
Chapter 2: Glide Overview ............................................................................................... 7
2.1 Introduction to Glide ................................................................................................. 7
2.2 Glide Constraints .................................................................................................... 12
2.3 Glide Extra-Precision Mode ................................................................................... 13
2.4 Glide/Prime Induced Fit Docking .......................................................................... 14
Chapter 3: Protein and Ligand Preparation........................................................ 15
3.1 Choosing the Most Appropriate Protein Site ..................................................... 15
3.2 Protein Preparation ................................................................................................. 16
3.3 Checking the Protein Structures .......................................................................... 19
3.3.1 Checking the Orientation of Water Molecules................................................... 19
3.3.2 Checking for Steric Clashes.............................................................................. 19
3.3.3 Resolving H-Bonding Conflicts ......................................................................... 19
3.3.4 Docking the Native Ligand ................................................................................ 20
3.4 Ligand Preparation.................................................................................................. 21
3.4.1 Using LigPrep for Ligand Preparation ............................................................... 22
3.4.2 Using Other Programs for Ligand Preparation.................................................. 24
Chapter 4: Receptor Grid Generation .................................................................... 25
4.1 The Receptor Grid Generation Panel ................................................................... 25
Glide 6.7 User Manual
iii
Contents
4.2 The Receptor Tab .................................................................................................... 27
4.2.1 Defining the Receptor ....................................................................................... 27
4.2.2 Global Van der Waals Radius Scaling .............................................................. 28
4.2.3 Selection of Partial Charges ............................................................................. 29
4.2.4 Grids for Peptide Docking ................................................................................. 29
4.2.5 Per-Atom Van der Waals Radius and Charge Scaling ...................................... 29
4.2.6 Hydrogen Bonds with Aromatic Hydrogens and Halogens ............................... 30
4.3 The Site Tab .............................................................................................................. 30
4.3.1 Selecting a Box Center ..................................................................................... 32
4.3.2 Setting the Box Sizes ........................................................................................ 33
4.4 The Constraints Tab ................................................................................................ 34
4.4.1 Setting Positional and NOE Constraints ........................................................... 36
4.4.2 Setting H-Bond and Metal Constraints.............................................................. 37
4.4.3 Setting Metal Coordination Constraints ............................................................ 39
4.5 The Rotatable Groups Tab ..................................................................................... 41
4.6 The Excluded Volumes Tab ................................................................................... 42
Chapter 5: Ligand Docking ............................................................................................ 45
5.1 The Ligand Docking Panel ..................................................................................... 45
5.2 Specifying the Receptor Grid ................................................................................ 47
5.3 The Ligands Tab ...................................................................................................... 47
5.3.1 Specifying the Source of the Ligands ............................................................... 47
5.3.2 Selecting the Source of Ligand Partial Charges ............................................... 48
5.3.3 Setting Restrictions on the Type of Ligands...................................................... 49
5.3.4 Van der Waals Radii Scaling............................................................................. 49
5.4 The Settings Tab ...................................................................................................... 50
5.4.1 Selecting the Docking Precision ....................................................................... 51
5.4.2 Setting Ligand Sampling Options...................................................................... 51
5.4.3 Writing XP Descriptor Information..................................................................... 53
5.4.4 Adding Epik State Penalties to the Docking Score ........................................... 53
5.4.5 Rewarding Intramolecular Hydrogen Bonds ..................................................... 54
iv
Schrödinger Software Release 2015-2
Contents
5.4.6 Improving Planarity Around sp2 Atoms ............................................................. 54
5.4.7 Applying Excluded Volumes.............................................................................. 54
5.4.8 Advanced Settings ............................................................................................ 54
5.4.8.1 Conformer Generation ........................................................................... 55
5.4.8.2 Selection of Initial Poses ........................................................................ 56
5.4.8.3 Energy Minimization Settings ................................................................. 57
5.4.8.4 Hydrogen Bonds with Aromatic Hydrogens and Halogens........................ 57
5.5 The Core Tab ............................................................................................................ 58
5.6 The Constraints Tab ................................................................................................ 60
5.6.1 Setting Constraints............................................................................................ 61
5.6.2 Defining Ligand Features .................................................................................. 63
5.6.2.1 Loading and Saving Feature Sets ........................................................... 65
5.6.2.2 Adding, Editing, and Deleting Patterns .................................................... 65
5.6.2.3 Excluding Functional Groups from a Feature ........................................... 66
5.6.2.4 Visualizing Patterns ............................................................................... 67
5.6.3 Choosing When To Apply Constraints .............................................................. 67
5.6.4 Using Multiple Constraints ................................................................................ 67
5.7 The Torsional Constraints Tab .............................................................................. 68
5.8 The Output Tab ........................................................................................................ 70
5.8.1 Structure Output Options .................................................................................. 70
5.8.2 Post-Docking Minimization ................................................................................ 72
5.8.3 Per-Residue Interaction Scores ........................................................................ 73
5.8.4 Advanced Settings ............................................................................................ 74
5.9 Docking Output Properties .................................................................................... 76
Chapter 6: Visualizing Docking Results ................................................................ 79
6.1 Viewing Poses .......................................................................................................... 79
6.1.1 Setting Up Pose Viewing................................................................................... 79
6.1.2 Stepping Through Poses................................................................................... 81
6.1.3 Displaying Nonbonded and Per-Residue Interactions....................................... 82
6.1.4 Exporting Poses for Rotatable Groups.............................................................. 83
Glide 6.7 User Manual
v
Contents
6.2 The Glide XP Visualizer .......................................................................................... 84
6.2.1 The Ligands Table............................................................................................. 85
6.2.2 Controlling the Display ...................................................................................... 87
6.2.3 Exporting Data and Structures.......................................................................... 88
6.2.4 XP Terms and Their Visualizations ................................................................... 88
Chapter 7: Running Glide from the Command Line ..................................... 93
7.1 Running Jobs ........................................................................................................... 94
7.2 Job Files and Directories ....................................................................................... 95
7.3 The glide Command ................................................................................................ 97
7.4 Restarting Glide Docking Jobs ........................................................................... 109
7.5 Glide Utilities .......................................................................................................... 109
7.5.1 glide_sort ........................................................................................................ 109
7.5.2 glide_merge .................................................................................................... 110
7.5.3 glide_rescore .................................................................................................. 110
7.6 Customizing Glide Calculations ......................................................................... 110
7.6.1 Changing the Glide Atom Typing .................................................................... 111
7.6.2 Defining Torsional Controls for Planar and Other Groups............................... 111
7.7 Docking Log Messages ........................................................................................ 113
Appendix A: Old Commands for Running Glide............................................ 115
A.1 The impact Command ......................................................................................... 115
A.2 The para_glide Command .................................................................................. 115
A.3 Recovering Failed Docking Jobs ...................................................................... 116
References .............................................................................................................................. 119
Getting Help ........................................................................................................................... 121
Index ............................................................................................................................................ 125
vi
Schrödinger Software Release 2015-2
Document Conventions
In addition to the use of italics for names of documents, the font conventions that are used in
this document are summarized in the table below.
Font
Example
Use
Sans serif
Project Table
Names of GUI features, such as panels, menus,
menu items, buttons, and labels
Monospace
$SCHRODINGER/maestro
File names, directory names, commands, environment variables, command input and output
Italic
filename
Text that the user must replace with a value
Sans serif
uppercase
CTRL+H
Keyboard keys
Links to other locations in the current document or to other PDF documents are colored like
this: Document Conventions.
In descriptions of command syntax, the following UNIX conventions are used: braces { }
enclose a choice of required items, square brackets [ ] enclose optional items, and the bar
symbol | separates items in a list from which one item must be chosen. Lines of command
syntax that wrap should be interpreted as a single command.
File name, path, and environment variable syntax is generally given with the UNIX conventions. To obtain the Windows conventions, replace the forward slash / with the backslash \ in
path or directory names, and replace the $ at the beginning of an environment variable with a %
at each end. For example, $SCHRODINGER/maestro becomes %SCHRODINGER%\maestro.
Keyboard references are given in the Windows convention by default, with Mac equivalents in
parentheses, for example CTRL+H (H). Where Mac equivalents are not given, COMMAND
should be read in place of CTRL. The convention CTRL-H is not used.
In this document, to type text means to type the required text in the specified location, and to
enter text means to type the required text, then press the ENTER key.
References to literature sources are given in square brackets, like this: [10].
Glide 6.7 User Manual
vii
viii
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 1
Chapter 1:
Introduction
The Glide User Manual is intended to help you perform ligand database screening and highaccuracy docking with Glide. Glide is run primarily from the Maestro graphical user interface,
but can also be run from the command line. Online help for Glide is available in Maestro,
although the information in this manual is generally more comprehensive.
Chapter 2 introduces the scientific methods and computational procedures used in Glide.
Chapter 3 describes the preparation of the protein and the ligands for use in Glide.
Chapter 4 describes the use of the Receptor Grid Generation panel to calculate the grids that
represent the receptor.
Chapter 5 describes the use of the Ligand Docking panel to set up and run docking jobs, and the
use of Glide constraints and distributed processing of multiple-ligand docking calculations.
Chapter 6 contains information on visualizing the results of Glide docking runs, using the Pose
View mode in the Project Table and the Glide XP Visualizer.
Chapter 7 contains information about running Glide, and its associated applications and utilities, from the command line.
The Glide Quick Start Guide contains tutorials intended to familiarize you with protein preparation, receptor grid generation, ligand docking, and viewing poses.
1.1
Running Schrödinger Software
Schrödinger applications can be run from a graphical interface or from the command line. The
software writes input and output files to a directory (folder) which is termed the working directory. If you run applications from the command line, the directory from which you run the
application is the working directory for the job.
Linux:
To run any Schrödinger program on a Linux platform, or start a Schrödinger job on a remote
host from a Linux platform, you must first set the SCHRODINGER environment variable to the
Glide 6.7 User Manual
1
Chapter 1: Introduction
installation directory for your Schrödinger software. To set this variable, enter the following
command at a shell prompt:
csh/tcsh:
setenv SCHRODINGER installation-directory
bash/ksh:
export SCHRODINGER=installation-directory
Once you have set the SCHRODINGER environment variable, you can run programs and utilities
with the following commands:
$SCHRODINGER/program &
$SCHRODINGER/utilities/utility &
You can start the Maestro interface with the following command:
$SCHRODINGER/maestro &
It is usually a good idea to change to the desired working directory before starting the Maestro
interface. This directory then becomes the working directory.
Windows:
The primary way of running Schrödinger applications on a Windows platform is from a graphical interface. To start the Maestro interface, double-click on the Maestro icon, on a Maestro
project, or on a structure file; or choose Start → All Programs → Schrodinger-2015-2 →
Maestro. You do not need to make any settings before starting Maestro or running programs.
The default working directory is the Schrodinger folder in your Documents folder.
If you want to run applications from the command line, you can do so in one of the shells that
are provided with the installation and have the Schrödinger environment set up:
• Schrödinger Command Prompt—DOS shell.
• Schrödinger Power Shell—Windows Power Shell (if available).
You can open these shells from Start → All Programs → Schrodinger-2015-2. You do not need
to include the path to a program or utility when you type the command to run it. If you want
access to Unix-style utilities (such as awk, grep, and sed), preface the commands with sh, or
type sh in either of these shells to start a Unix-style shell.
Mac:
The primary way of running Schrödinger software on a Mac is from a graphical interface. To
start the Maestro interface, click its icon on the dock. If there is no Maestro icon on the dock,
you can put one there by dragging it from the SchrodingerSuite2015-2 folder in your Applications folder. This folder contains icons for all the available interfaces. The default working
2
Schrödinger Software Release 2015-2
Chapter 1: Introduction
directory is the Schrodinger folder in your Documents folder ($HOME/Documents/
Schrodinger).
Running software from the command line is similar to Linux—open a terminal window and
run the program. You can also start Maestro from the command line in the same way as on
Linux. The default working directory is then the directory from which you start Maestro. You
do not need to set the SCHRODINGER environment variable, as this is set in your default environment on installation. To set other variables, on OS X 10.7 use the command
defaults write ~/.MacOSX/environment variable "value"
and on OS X 10.8, 10.9, and 10.10 use the command
launchctl setenv variable "value"
1.2
Starting Jobs from the Maestro Interface
To run a job from the Maestro interface, you open a panel from one of the menus (e.g. Tasks),
make settings, and then submit the job to a host or a queueing system for execution. The panel
settings are described in the help topics and in the user manuals. When you have finished
making settings, you can use the Job toolbar to start the job.
You can start a job immediately by clicking Run. The job is run on the currently selected host
with the current job settings and the job name in the Job name text box. If you want to change
the job name, you can edit it in the text box before starting the job. Details of the job settings
are reported in the status bar, which is below the Job toolbar.
If you want to change the job settings, such as the host on which to run the job and the number
of processors to use, click the Settings button. (You can also click the arrow next to the button
and choose Job Settings from the menu that is displayed.)
You can then make the settings in the Job Settings dialog box, and choose to just save the
settings by clicking OK, or save the settings and start the job by clicking Run. These settings
apply only to jobs that are started from the current panel.
If you want to save the input files for the job but not run it, click the Settings button and choose
Write. A dialog box opens in which you can provide the job name, which is used to name the
files. The files are written to the current working directory.
Glide 6.7 User Manual
3
Chapter 1: Introduction
The Settings button also allows you to change the panel settings. You can choose Read, to read
settings from an input file for the job and apply them to the panel, or you can choose Reset
Panel to reset all the panel settings to their default values.
You can also set preferences for all jobs and how the interface interacts with the job at various
stages. This is done in the Preferences panel, which you can open at the Jobs section by
choosing Preferences from the Settings button menu.
Note:
The items present on the Settings menu can vary with the application. The descriptions
above cover all of the items.
The icon on the Job Status button shows the status of jobs for the application that belong to the
current project. It starts spinning when the first job is successfully launched, and stops spinning
when the last job finishes. It changes to an exclamation point if a job is not launched successfully.
Clicking the button shows a small job status window that lists the job name and status for all
active jobs submitted for the application from the current project, and a summary message at
the bottom. The rows are colored according to the status: yellow for submitted, green for
launched, running, or finished, red for incorporated, died, or killed. You can double-click on a
row to open the Monitor panel and monitor the job, or click the Monitor button to open the
Monitor panel and close the job status window. The job status is updated while the window is
open. If a job finishes while the window is open, the job remains displayed but with the new
status. Click anywhere outside the window to close it.
Jobs are run under the Job Control facility, which manages the details of starting the job, transferring files, checking on status, and so on. For more information about this facility and how it
operates, as well as details of the Job Settings dialog box, see the Job Control Guide.
1.3
Using Glide
The Glide task most frequently performed is ligand docking. The grid files produced by a
single receptor grid generation task can be used for any number of jobs that dock ligands to
that receptor. Before generating receptor grids, it is strongly recommended that you prepare the
protein. Therefore, the first steps of a typical project beginning with an unprepared proteinligand complex structure (e.g., from PDB) might proceed using the Glide panels as follows:
1. Prepare the receptor using the Protein Preparation Wizard panel. See Chapter 3 for
details.
2. Ensure that the ligands to be docked are in the right form. See Chapter 3 for details.
4
Schrödinger Software Release 2015-2
Chapter 1: Introduction
3. With the prepared receptor-ligand complex in the Workspace, use the Receptor Grid Generation panel to specify settings, optionally define constraints, and start the receptor grid
generation job. Details of setting up receptor grid generation jobs are given in Chapter 4.
4. Specify the base name for the receptor grid files you want to use in the Ligand Docking
panel, and use the other settings and options in the panel to set up and start a ligand docking job. As many docking jobs as you want can be set up in this panel, using the current
receptor grids or specifying a different set of grids to use. Details of setting up ligand
docking jobs are given in Chapter 5.
1.4
Documentation
For information related to the installation and use of Glide, see the following documentation:
• The Installation Guide, which includes installation instructions for all Schrödinger products and documentation.
• The Job Control Guide, which includes instructions for running and managing
Schrödinger jobs.
• The Glide Quick Start Guide, which contains tutorials intended to familiarize you with
receptor grid generation, ligand docking, and visualization of results.
• The Impact Command Reference Manual, which contains syntax and keywords for
Impact command input files.
• The Maestro User Manual, which describes how to use the features of Maestro, including
the Atom Selection dialog box.
• The Maestro Command Reference Manual, which contains commands, options, and arguments for running Maestro from the command line, including the Atom Specification
Language (ASL) and the Entry Specification Language (ESL).
1.5
Citing Glide in Publications
The use of this product should be acknowledged in publications as:
Glide, version 6.7, Schrödinger, LLC, New York, NY, 2015.
In addition, please cite the following papers:
• Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.;
Repasky, M. P.; Knoll, E. H.; Shaw, D. E.; Shelley, M.; Perry, J. K.; Francis, P.; Shenkin,
P. S. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and
Assessment of Docking Accuracy, J. Med. Chem. 2004, 47, 1739-1749.
Glide 6.7 User Manual
5
Chapter 1: Introduction
• Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye, L. L.; Pollard, W. T.;
Banks, J. L. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 2.
Enrichment Factors in Database Screening. J. Med. Chem. 2004, 47, 1750–1759.
• Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.; Greenwood, J. R.; Halgren, T.
A.; Sanschagrin, P. C.; Mainz, D. T. Extra Precision Glide: Docking and Scoring Incorporating a Model of Hydrophobic Enclosure for Protein-Ligand Complexes. J. Med. Chem.
2006, 49, 6177–6196.
6
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 2
Chapter 2:
Glide Overview
This chapter contains an overview of the Glide (Grid-based Ligand Docking with Energetics)
program, its scientific methods and computational procedures.
Glide searches for favorable interactions between one or more ligand molecules and a receptor
molecule, usually a protein. Each ligand must be a single molecule, while the receptor may
include more than one molecule, e.g., a protein and a cofactor. Glide can be run in rigid or flexible docking modes; the latter automatically generates conformations for each input ligand.
The combination of position and orientation of a ligand relative to the receptor, along with its
conformation in flexible docking, is referred to as a ligand pose.
The ligand poses that Glide generates pass through a series of hierarchical filters that evaluate
the ligand’s interaction with the receptor. The initial filters test the spatial fit of the ligand to the
defined active site, and examine the complementarity of ligand-receptor interactions using a
grid-based method patterned after the empirical ChemScore function [1].
Poses that pass these initial screens enter the final stage of the algorithm, which involves evaluation and minimization of a grid approximation to the OPLS-AA nonbonded ligand-receptor
interaction energy.
Final scoring is then carried out on the energy-minimized poses. By default, Schrödinger’s
proprietary GlideScore multi-ligand scoring function is used to score the poses. If GlideScore
was selected as the scoring function, a composite Emodel score is then used to rank the poses
of each ligand and to select the poses to be reported to the user. Emodel combines GlideScore,
the nonbonded interaction energy, and, for flexible docking, the excess internal energy of the
generated ligand conformation.
2.1
Introduction to Glide
Glide uses a hierarchical series of filters to search for possible locations of the ligand in the
active-site region of the receptor. The shape and properties of the receptor are represented on a
grid by several different sets of fields that provide progressively more accurate scoring of the
ligand poses.
Conformational flexibility is handled in Glide by an extensive conformational search,
augmented by a heuristic screen that rapidly eliminates unsuitable conformations, such as
conformations that have long-range internal hydrogen bonds.
Glide 6.7 User Manual
7
Chapter 2: Glide Overview
H
Rotamer group
N
S
O
N
O
Rotamer group
O
Figure 2.1. Definition of core and rotamer groups.
As illustrated in Figure 2.1, each ligand is divided into a core region and some number of
rotamer groups. Each rotamer group is attached to the core by a rotatable bond, but does not
contain additional rotatable bonds. The core is what remains when each terminus of the ligand
is severed at the “last” rotatable bond. Carbon and nitrogen end groups terminated with
hydrogen (—CH3, —NH2, —NH3+) are not considered rotatable because their conformational
variation is of little significance. In Figure 2.1, the four central torsions are part of the core, and
the methyl groups are not considered rotatable.
During conformation generation, each core region is represented by a set of core conformations, the number of which depends on the number of rotatable bonds, conformationally labile
5– and 6–membered rings, and asymmetric pyramidal trigonal nitrogen centers in the core.
This set typically contains fewer than 500 core conformations, even for quite large and flexible
ligands, and far fewer for more rigid ligands. Every rotamer state for each rotamer group
attached to the core is enumerated. The core plus all possible rotamer-group conformations is
docked as a single object. Glide can also dock sets of pre-computed conformations. However,
Glide offers its greatest value when conformations are generated internally.
For each core conformation (or for rigid docking, each ligand), an exhaustive search of
possible locations and orientations is performed over the active site of the protein. The search
begins with the selection of “site points” on an equally spaced 2 Å grid that covers the activesite region (Stage 1 in Figure 2.2). This selection is made as follows. Distances from the site
point to the receptor surface are evaluated at a series of pre-specified directions and sorted into
8
Schrödinger Software Release 2015-2
Chapter 2: Glide Overview
distance ranges (“bins”) of width 1 Å. Likewise, distances from the ligand center (the midpoint
of the two most widely separated atoms) to the ligand surface are sorted into bins of width 1 Å.
For a given site point, the distance ranges from the site point to the receptor are compared with
those from the ligand center to the ligand surface. Glide positions the ligand center at the site
point if there is a good enough match, but skips over the site point if there is not.
The second stage of the hierarchy begins by examining the placement of atoms that lie within a
specified distance of the line drawn between the most widely separated atoms (the ligand
diameter). This is done for a pre-specified selection of possible orientations of the ligand diameter (Step 2a). If there are too many steric clashes with the receptor, the orientation is skipped.
Glide “Funnel”
Ligand conformations
Stage 1. Site-point search
Stage 2: Step 2a. Diameter test
Step 2b. Subset test
Step 2c. Greedy score
Step 2d. Refinement
Stage 3. Grid minimization
Stage 4. Final scoring
(GlideScore)
Top hits (structure and report files)
Figure 2.2. The Glide docking hierarchy.
Glide 6.7 User Manual
9
Chapter 2: Glide Overview
Next (Step 2b), rotation about the ligand diameter is considered, and the interactions of a
subset consisting of all atoms capable of making hydrogen bonds or ligand-metal interactions
with the receptor are scored (subset test). If this score is good enough, all interactions with the
receptor are scored (Step 2c).
The scoring in these three tests is carried out using Schrödinger’s discretized version of the
ChemScore empirical scoring function [1]. Much as for ChemScore itself, this algorithm
recognizes favorable hydrophobic, hydrogen-bonding, and metal-ligation interactions, and
penalizes steric clashes. This stage is called “greedy scoring,” because the actual score for each
atom depends not only on its position relative to the receptor but also on the best possible score
it could get by moving ±1 Å in x, y, or z. This is done to mute the sting of the large 2 Å jumps
in the site-point/ligand-center positions. The final step in Stage 2 is to re-score the top greedyscoring poses via a “refinement” procedure (Step 2d), in which the ligand as a whole is allowed
to move rigidly by ±1 Å in the Cartesian directions.
Only a small number of the best refined poses (typically 100-400) is passed on to the third
stage in the hierarchy—energy minimization on the pre-computed OPLS van der Waals and
electrostatic grids for the receptor. The energy minimization typically begins on a set of van
der Waals and electrostatic grids that have been “smoothed” to reduce the large energy and
gradient terms that result from too-close interatomic contacts. It finishes on the full-scale
OPLS nonbonded energy surface (“annealing”). This energy minimization consists only of
rigid-body translations and rotations when external conformations are docked. When conformations are generated internally, however, the optimization also includes torsional motion
about the core and end-group rotatable bonds. Unless otherwise specified, a small number of
the top-ranked poses are then subjected to a sampling procedure in which alternative localminima core and rotamer-group torsion angles are examined to try to improve the energy score.
Finally, the minimized poses are re-scored using Schrödinger’s proprietary GlideScore scoring
function. GlideScore is based on ChemScore, but includes a steric-clash term, adds other
rewards and penalties such as buried polar terms (devised by Schrödinger to penalize electrostatic mismatches), amide twist penalties, hydrophobic enclosure terms, and excluded volume
penalties, and has modifications to other terms:
GScore = 0.05*vdW + 0.15*Coul + Lipo + Hbond + Metal + Rewards + RotB + Site
The components of the GlideScore (GScore) are described in Table 2.1.
The choice of best-docked structure for each ligand is made using a model energy score
(Emodel) that combines the energy grid score, the binding affinity predicted by GlideScore,
and (for flexible docking) the internal strain energy for the model potential used to direct the
conformational-search algorithm. Glide also computes a specially constructed Coulomb-van
der Waals interaction-energy score (CvdW) that is formulated to avoid overly rewarding
charge-charge interactions at the expense of charge-dipole and dipole-dipole interactions. This
10
Schrödinger Software Release 2015-2
Chapter 2: Glide Overview
Table 2.1. GlideScore components.
Component
Description
vdW
Van der Waals energy. This term is calculated with reduced net ionic charges on
groups with formal charges, such as metals, carboxylates, and guanidiniums.
Coul
Coulomb energy. This term is calculated with reduced net ionic charges on groups
with formal charges, such as metals, carboxylates, and guanidiniums.
Lipo
Lipophilic term, which is a pairwise term in SP but is derived from the hydrophobic grid potential for XP. Rewards favorable hydrophobic interactions.
HBond
Hydrogen-bonding term. This term is separated into differently weighted components that depend on whether the donor and acceptor are neutral, one is neutral and
the other is charged, or both are charged.
Metal
Metal-binding term. Only the interactions with anionic or highly polar acceptor
atoms are included. If the net metal charge in the apo protein is positive, the preference for anionic or polar ligands is included; if the net charge is zero, the preference is suppressed.
Rewards
Rewards and penalties for various features, such as buried polar groups, hydrophobic enclosure, correlated hydrogen bonds, amide twists, and so on. This category
covers all terms other than those explicitly mentioned.
RotB
Penalty for freezing rotatable bonds.
Site
Polar interactions in the active site. Polar but non-hydrogen-bonding atoms in a
hydrophobic region are rewarded.
score is intended to be more suitable for comparing the binding affinities of different ligands
than is the “raw” Coulomb-van der Waals interaction energy. In addition to the GlideScore, a
“docking score” is reported, which is the GlideScore supplemented by Epik state penalties, if
used, and strain corrections, if used.
This hierarchical search gives Glide exceptionally high accuracy in predicting the binding
mode of the ligand. At the same time, the computational cost is dramatically reduced compared
to what would be required for a complete systematic search. The key to this reduction is that
the algorithm allows the rotamer groups to be optimized one at a time for a given core conformation and location of the ligand. For example, if there are five rotamer groups and each has
three rotamer states, the total number of conformers in the ensemble based on this core conformation/location is 35 = 243. However, if the rotamer groups are optimized one at a time, the
number of conformational combinations is only 3×5 = 15, for a savings of about a factor of 15
in computational effort. While many other time-saving algorithms in Glide contribute to its
performance advantages, this fundamental qualitative feature allows large libraries to be
screened at an affordable computational cost.
Glide 6.7 User Manual
11
Chapter 2: Glide Overview
2.2
Glide Constraints
A Glide constraint is a ligand-receptor interaction requirement. The constraint usually means
that a ligand atom must lie within a certain region defined in relation to features of the receptor
that are responsible for ligand binding. To use Glide constraints, you must specify receptor
sites for possible ligand interactions when you set up a receptor grid generation job. When you
run a ligand docking job, you can select Glide constraints to apply from the list of receptor
constraint sites that you defined for the receptor.
Other constraints that reflect ligand-receptor interactions indirectly are applied during docking.
These include superposition of the ligand core on a reference core, and restriction of specified
ligand torsions.
In Glide constraint docking jobs, Glide incorporates satisfaction of these constraints into
several of its hierarchical filters, allowing prompt rejection of docked poses that fail to meet the
requirements.
The first constraint filter is a simple one involving the atoms of the ligand. If a ligand does not
contain atoms of the right types to make the required interactions with the receptor constraint
atoms, Glide simply skips that ligand—for instance if a selected receptor constraint atom is a
polar hydrogen and the ligand has no hydrogen-bond acceptors. If there are qualifying atoms in
the ligand, Glide keeps a list of the possible “partner atoms” for each constraint, for use in
subsequent filters. For hydrophobic constraints, this filter checks that the ligand contains a
sufficient number of hydrophobic heavy atoms; the sum of the minimum numbers specified for
all the selected constraints of this type. In addition, Glide locates centroids of ligand hydrophobic groups, which take the role of partner atoms in subsequent filters.
Other filters operate further down the funnel. These include matching distances between
partner atoms (or hydrophobic group centroids) for different constraints against the corresponding distances between receptor atoms; matching distances from the ligand center to
partner atoms against those from receptor site points to constraint atoms, and restricting orientations of, and rotations about, the ligand diameter to those that bring partner atoms into appropriate proximity with receptor atoms or regions. In addition, a restraining potential is used in
grid optimizations (for hydrogen-bond and metal constraints only) to bring or keep constraintsatisfying ligand-receptor atom pairs at appropriate distances, and torsional sampling moves
are not tried if they would move any ligand atom currently in a constraint-satisfying position.
Finally, grid-optimized poses are rejected if they don’t strictly satisfy all selected constraints.
Core constraints are applied by a “snap and refine” algorithm. The core pattern is identified in
the ligand, and the matching atoms are placed at exactly the same coordinates as the reference
ligand. The rest of the molecule is rebuilt, preserving the internal coordinates of the non-core
atoms. The rotatable bonds are then sampled, and passed through rough scoring, but the refine-
12
Schrödinger Software Release 2015-2
Chapter 2: Glide Overview
ment part is skipped, as the core is placed exactly and no translation of the ligand is needed.
The refinement (grid minimization) is performed, in which the non-core torsions are minimized, and sampling of torsions is then applied to improve poses, if possible. Finally, postdocking minimization is performed, with steep, flat-bottomed constraints on the core atoms.
This last step may change the RMSD with respect to the reference core; prior to that, it is zero.
For information on using Glide constraints, see Section 4.4, Section 4.6, Section 5.5,
Section 5.6, and Section 5.7.
2.3
Glide Extra-Precision Mode
The extra-precision (XP) mode of Glide combines a powerful sampling protocol with the use
of a custom scoring function designed to identify ligand poses that would be expected to have
unfavorable energies, based on well-known principles of physical chemistry. The presumption
is that only active compounds will have available poses that avoid these penalties and also
receive favorable scores for appropriate hydrophobic contact between the protein and the
ligand, hydrogen-bonding interactions, and so on. The chief purposes of the XP method are to
weed out false positives and to provide a better correlation between good poses and good
scores.
Extra-precision mode is a refinement tool designed for use only on good ligand poses. The
more extensive XP docking method and specialized XP scoring method are strongly coupled:
the more precise poses produced by XP docking are necessary for the more demanding XP
scoring method. Because XP docking mode requires considerably more CPU time, you should
screen large sets of ligands first in standard-precision (SP) mode or in high-throughput virtual
screening (HTVS) mode. Only the top-scoring ligands should be docked using XP mode.
In any grid-based docking method, the receptor is essentially frozen. Some degree of flexibility
can be introduced by scaling parts of the potential, as is done in SP mode, but such techniques
cannot represent larger adjustments to the receptor, like conformational changes. XP mode is
less forgiving than SP mode so that it can screen out false positives, and is designed to locate
active compounds that bind to a particular conformation of the receptor. Active compounds can
be prevented from docking if these compounds are not compatible with the particular conformation of the receptor that is being used. To ensure that actives are not eliminated you should if
possible dock into multiple receptor conformations and combine the results of the individual
docking runs.
The XP sampling method is based on an anchor and refined growth strategy. Anchor fragments
of the docked ligand, typically rings, are chosen from the set of SP poses and the molecule is
re-grown bond by bond from these anchor positions. Complete minimizations and XP scoring
are carried out on the large ensemble of poses generated by this growing method. At various
cycles the growing is focused to attempt to relieve any XP scoring penalties as well as to optiGlide 6.7 User Manual
13
Chapter 2: Glide Overview
mize the best scoring poses. This focused sampling is essential for allowing the use of the hard
XP scoring function as well as for finding the best scoring basins of attraction. It is important
to note that the coupling between the extra sampling and the XP scoring means that it is not
recommended to just score the SP poses with XP scoring.
The scoring function (GlideScore XP) includes additional terms over the SP scoring function,
and a more complete treatment of some of the SP terms, as described below.
To model solvation, explicit water molecules are docked into a list of protein-ligand complexes
that otherwise receive good GlideScores, and descriptors based on the interaction of these
water molecules with various charged and polar groups of the ligand and protein are used as a
measure of whether the complex is physically realistic. Penalties are assigned to structures
where statistical results indicate that one or more groups is inadequately solvated. A large database of co-crystallized structures has been used to optimize the parameters associated with the
penalty terms. The explicit-water technology and descriptors are also used in Glide SP scoring.
However, the improved sampling allows XP docking to assign substantially higher penalties to
serious violations of physical principles.
GlideScore XP specifically rewards occupancy of well-defined hydrophobic pockets by hydrophobic ligand groups. Hydrophobic reward terms are employed in empirical scoring functions
such as ChemScore and the SP version of GlideScore in the form of lipophilic-lipophilic pair
terms, while other empirical scoring functions use lipophilic surface-area contact terms for this
purpose. Investigations have shown that simple pair terms underestimate hydrophobic effects
in certain well-defined cases. The hydrophobic term in GlideScore XP was developed to offset
this underestimation. The term can confer up to several kcal/mol of additional binding energy
in favorable cases, and substantially improves enrichment factors GlideScore XP also includes
improvements to the scoring of hydrogen bonds as well as detection of buried polar groups,
and detection of pi-cation and pi-pi stacking interactions.
For information on using XP mode, see Section 5.4. For information on visualizing the various
XP scoring terms, see Section 6.2.
2.4
Glide/Prime Induced Fit Docking
Glide docking uses the basic assumption of a rigid receptor. Scaling of van der Waals radii of
nonpolar atoms, which decreases penalties for close contacts, can be used to model a slight
“give” in the receptor or the ligand or both, and specified hydroxyl groups can be allowed to
reorient to optimize hydrogen bonding. This may not be sufficient to treat systems where
ligand binding induces substantial conformation changes in the receptor (“induced fit”).
Schrödinger has developed a procedure for such cases which uses Prime and Glide to perform
induced fit docking. For more information about Schrödinger’s Induced Fit Docking protocol,
see the document Induced Fit Docking.
14
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 3
Chapter 3:
Protein and Ligand Preparation
The quality of Glide results depends on reasonable starting structures for both the protein and
the ligand. Schrödinger offers a comprehensive protein preparation facility in the Protein Preparation Wizard, which is designed to ensure chemical correctness and to optimize protein
structures for use with Glide and other products. Likewise, Schrödinger offers a comprehensive ligand preparation facility in LigPrep. It is strongly recommended that you process protein
and ligand structures with these facilities in order to achieve the best results.
3.1
Choosing the Most Appropriate Protein Site
Glide relies on the rigid-receptor approximation in order to treat protein-ligand binding.
However, protein conformations are in general flexible and can occupy a continuum of states.
Proteins can exhibit induced fit effects on binding of a ligand, in which the protein conformation changes significantly. This effect can be seen in cases where more that one co-crystallized
complex is available. Since Glide docking experiments employ only a single protein geometry,
two approaches are commonly taken. One is to select a single well-suited representative
protein structure to dock into. The other is to use an ensemble of representative structures, into
which each of the candidate ligands is docked.
If more than one co-crystallized complex is available, you must decide whether to select a
single protein site or to choose two or more sites for use in independent docking experiments.
This choice may depend on whether the protein site is rigid or mobile, as well as on the
resources available and the objectives of the docking study. A single representative site should
suffice for a rigid protein. Cases in which the site changes substantially as different ligands
bind may require the use of two or more sites, if finding the maximum number of promising
ligands is the main objective.
Some proteins are known to be rigid. To determine whether a single site is likely to be sufficient in other cases, you can transpose the known protein-ligand complexes into the coordinate
frame of a “reference” complex. The objective is to judge whether the reference site is compatible with all the co-crystallized ligands or whether another site appears more suitable. A procedure for making this determination is as follows:
1. Choose a reference complex and superimpose all the other complexes to it.
You can perform this task in Maestro with the Superposition panel. The atoms you
choose to superimpose could be the alpha carbons, or all the backbone atoms for residues
Glide 6.7 User Manual
15
Chapter 3: Protein and Ligand Preparation
in common, or the Cα or backbone atoms for selected active-site residues. Alternatively,
you can use the Protein Structure Alignment facility to perform the superposition in an
automated fashion.
2. Display the protein for the reference complex and the ligand for each of the other complexes, in turn. Examine the active-site region to determine whether the superimposed
ligand can fit into the reference site without steric clashes that could not reasonably be
relieved by minor repositioning. You can use the tools on the Measurements toolbar or in
the Non-Bonded Interactions panel to examine contacts.
3. Display the protein for the reference complex and the protein for each of the other complexes in turn. Note whether any residues in the superimposed protein differ appreciably
in position or conformation from those in the reference site.
4. From the above steps, judge whether the reference site appears compatible with all the
co-crystallized ligands or, if not, whether another site appears more compatible.
5. Choose a most representative (i.e., fairly generous) site for docking—or choose two or
more sites if there are large differences between the sites and the objective is to find as
many prospective strong binders as possible.
6. Write out a separate file for the protein or proteins that will be prepared. Also write out
separate files for the superimposed ligands.
For example, an initial screening of the CDK-2 kinase receptor used 1hck as the docking site,
the co-crystallized ligand for which is ATP. Other known co-crystallized complexes include
1aq1, 1di8, 1dm2, 1fvt, and 1fvv. Superimposing these five complexes onto 1hck using all Cα
atoms in common revealed that at least four of the five ligands (all active binders) cannot fit
into the 1hck site because its active-site channel is too short. The reason is that ATP and a Mg2+
ion bound to its terminal phosphate group pull glutamate and lysine sidechains more deeply
into the 1hck cavity, where they form an ion pair that closes off the cleft. Based on this visual
examination, the 1dm2 site was chosen instead. This site is considerably more open than the
1hck site, though somewhat less so than the 1aq1 or 1fvv sites. Glide was far more successful
in docking the known binders into the 1dm2 site than into 1hck, and was even more successful
when the still more open 1aq1 site was used.
3.2
Protein Preparation
A typical PDB structure file consists only of heavy atoms, can contain waters, cofactors, and
metal ions, and can be multimeric. The structure generally has no information on bond orders,
topologies, or formal atomic charges. Terminal amide groups can also be misaligned, because
the X-ray structure analysis cannot usually distinguish between O and NH2. Ionization and
tautomeric states are also generally unassigned. Glide calculations use an all-atom force field
16
Schrödinger Software Release 2015-2
Chapter 3: Protein and Ligand Preparation
for accurate energy evaluation. Thus, Glide requires bond orders and ionization states to be
properly assigned and performs better when side chains are reoriented when necessary and
steric clashes are relieved.
This section provides an overview of the protein preparation process. The entire procedure can
be performed in the Protein Preparation Wizard panel, which you open from the Workflows
menu on the main toolbar. This tool and its use is described in detail in Chapter 2 of the Protein
Preparation Guide.
After processing, you will have files containing refined, hydrogenated structures of the ligand
and the ligand-receptor complex. The prepared structures are suitable for use with Glide. In
most cases, not all of the steps (outlined below) need to be performed. See the descriptions of
each step to determine whether it is required, and make the appropriate selections in the Protein
Preparation Wizard panel.
You may on occasion want to perform some of these steps manually. Detailed procedures for
manual preparation are described in Chapter 3 of the Protein Preparation Guide.
The steps in the procedure, which can be performed either in the Protein Preparation Wizard
panel or manually, are listed below.
1. Import a ligand/protein cocrystallized structure, typically from PDB, into Maestro.
The preparation component of the protein preparation facility requires an identified
ligand.
2. Simplify multimeric complexes.
For computational efficiency it is desirable to keep the number of atoms in the complex
structure to a minimum. If the binding interaction of interest takes place within a single
subunit, you should retain only one ligand-receptor subunit to prepare for Glide. If two
identical chains are both required to form the active site, neither should be deleted.
• Determine whether the protein-ligand complex is a dimer or other multimer containing duplicate binding sites and duplicate chains that are redundant.
• If the structure is a multimer with duplicate binding sites, remove redundant binding
sites and the associated chains by picking and deleting molecules or chains.
3. Locate any waters you want to keep, then delete all others.
These waters are identified by the oxygen atom, and usually do not have hydrogens
attached. Generally, all waters (except those coordinated to metals) are deleted, but
waters that bridge between the ligand and the protein are sometimes retained. If waters
are kept, hydrogens will be added to them by the preparation component of the protein
preparation job. Afterwards, check that these water molecules are correctly oriented.
Glide 6.7 User Manual
17
Chapter 3: Protein and Ligand Preparation
4. Adjust the protein, metal ions, and cofactors.
Problems in the PDB protein structure may need to be repaired before it can be used.
Incomplete residues are the most common errors, but may be relatively harmless if they
are distant from the active site. Structures that are missing residues near the active site
should be repaired.
Covalent bonds from metal ions to the protein should be changed to zero-order bonds,
and the formal charges on the metal and the ligating groups should be adjusted to appropriate values.
Cofactors are included as part of the protein, but because they are not standard residues it
is sometimes necessary to use Maestro’s structure-editing capabilities to ensure that multiple bonds and formal charges are assigned correctly.
• Fix any serious errors in the protein.
• Check the protein structure for metal ions and cofactors.
• If there are bonds to metal ions, change the bonds to zero-order bonds and adjust the
formal charges of the metal and the atoms that were attached to it.
• Set charges and correct atom types for any metal atoms, as needed.
• Set bond orders and formal charges for any cofactors, as needed.
5. Adjust the ligand bond orders and formal charges.
If the complex structure contains bonds from the ligand or a cofactor to a protein metal,
they must be deleted. Glide models such interactions as van der Waals plus electrostatic
interactions. Glide cannot handle normal covalent bonds to the ligand, such as might be
found in an acyl enzyme.
If you are working with a dimeric or large protein and two ligands exist in two active
sites, the bond orders have to be corrected in both ligand structures.
6. Run a restrained minimization of the protein structure.
This is done with impref, and should reorient side-chain hydroxyl groups and alleviate
potential steric clashes. The minimization is restrained to the input protein coordinates by
a user-selected RMSD tolerance.
7. Review the prepared structures.
• If problems arise during the restrained minimization, review the log file, correct the
problems, and rerun.
• Examine the refined ligand/protein/water structure for correct formal charges, bond
orders, and protonation states and make final adjustments as needed.
18
Schrödinger Software Release 2015-2
Chapter 3: Protein and Ligand Preparation
3.3
Checking the Protein Structures
After you have completed the protein preparation, you should check the completed ligand and
protein structures.
3.3.1
Checking the Orientation of Water Molecules
You only need to perform this step if you kept some structural waters. Reorienting the hydrogens is not strictly necessary, as their orientation should have been changed during refinement,
but it is useful to check that the orientation is correct.
If the orientation is incorrect, reorient the molecules by using the procedure outlined in
Section 3.9 of the Protein Preparation Guide.
When you have corrected the orientation of the retained water molecules, you should run a
refinement on the adjusted protein-ligand complex.
3.3.2
Checking for Steric Clashes
You should make sure that the prepared site accommodates the co-crystallized ligand in the
restraint-optimized geometry obtained from the structure preparation.
Steric clashes can be detected by displaying the ligand and protein in Maestro and using the
Measurements toolbar or the Non-Bonded Interactions panel to visualize bad or ugly contacts.
Maestro defines bad contacts purely on the basis of the ratio of the interatomic distance to the
sum of the van der Waals radii it assigns. As a result, normal hydrogen bonds are classified as
bad or ugly contacts. By default, Maestro filters out contacts that are identified as hydrogen
bonds, and displays only the genuine bad or ugly contacts.
If steric clashes are found, repeat the restrained optimization portion of the protein preparation
procedure, but allow a greater rms deviation from the starting heavy-atom coordinates than the
default of 0.3 Å. Alternatively, you can apply an additional series of restrained optimizations to
the prepared ligand-protein complex to allow the site to relax from its current geometry.
3.3.3
Resolving H-Bonding Conflicts
You should look for inconsistencies in hydrogen bonding to see whether a misprotonation of
the ligand or the protein might have left two acceptor atoms close to one another without an
intervening hydrogen bond. One or more residues may need to be modified to resolve such an
acceptor-acceptor or donor-donor clash.
Some of these clashes are recognized by the preparation process but cannot be resolved by it.
The preparation process may have no control over other clashes. An example of the latter typi-
Glide 6.7 User Manual
19
Chapter 3: Protein and Ligand Preparation
cally occurs in an aspartyl protease such as HIV, where both active-site aspartates are close to
one or more atoms of a properly docked ligand. Because these contact distances fall within any
reasonable cavity radius, the carboxylates are not subject to being neutralized and will both be
represented as negatively charged by the preparation process. However, when the ligand interacts with the aspartates via a hydroxyl group or similar neutral functionality, one of the aspartates is typically modeled as neutral.
If residues need to be modified, follow these steps:
1. Place the refined protein-ligand complex in the Workspace.
2. Examine the interaction between the ligand and the protein (and/or the cofactor).
3. Use your judgment and chemical intuition to determine which protonation state and tautomeric form the residues in question should have.
4. Use the structure-editing capabilities in Maestro to resolve the conflict (see Section 3.8 of
the Protein Preparation Guide for procedures).
5. Re-minimize the structure.
It is usually sufficient to add the proton and perform about 50 steps of steepest-descent minimization to correct the nearby bond lengths and angles. Because this optimizer does not make
large-scale changes, the partial minimization can be done even on the isolated ligand or protein
without danger of altering the conformation significantly. However, if comparison to the original complex shows that the electrostatic mismatch due to the misprotonation has appreciably
changed the positions of the ligand or protein atoms during the protein-preparation procedure,
it is best to reprotonate the original structure and redo the restrained minimization.
3.3.4
Docking the Native Ligand
Once you have prepared the protein and generated grids (see Chapter 4), you should dock the
native ligand both rigidly and flexibly. If either run fails to produce a low-rms structure, the
structure or structures obtained from the docking should be scored in place, and the outcomes
in the *.rept files for the doc king runs and in the *.scor files for the score-in-place runs
should be checked.
Before running a docking job, you can run a score-in-place Glide calculation on the complex
and check that the metal-ligation energy is reasonable. If it is highly positive, you may have to
re-adjust the charge and protonation states in the active site manually.
• If rigid docking fails to give a low-rms pose, there may be a problem with the structure.
Double-check the protein preparation. Such a failure may also reflect incomplete sampling by Glide. If the problem is sampling, there are some settings in the GUI, and others
that can be made in the input file, that can be used to increase the amount of sampling.
20
Schrödinger Software Release 2015-2
Chapter 3: Protein and Ligand Preparation
• If rigid docking succeeds but flexible docking gives a high-rms pose that has a poor Coulomb-vdW energy, there may also be a sampling problem.
• If rigid docking succeeds but flexible docking finds a high-rms solution that has a more
negative Coulomb-vdW energy than the ligand pose generated in the structure preparation, there may be a scoring problem, that is, Glide may have chosen the wrong solution
because its Coulomb-vdW and Emodel energies are more favorable than those for the
correct solution. There usually is no way you can fix such a problem. However, if the GlideScore value computed for the misdocked ligand is similar to that obtained for the cocrystallized ligand, the database screen may still be effective in identifying ligands that
can bind tightly. The ligand may just have found a plausible alternative binding mode.
3.4
Ligand Preparation
To give the best results, the structures that are docked must be good representations of the
actual ligand structures as they would appear in a protein-ligand complex. This means that the
structures supplied to Glide must meet the following conditions:
1. They must be three-dimensional (3D).
2. They must have realistic bond lengths and bond angles.
Glide only modifies the torsional internal coordinates of the ligand during docking, so the
rest of the geometric parameters must be optimized beforehand.
3. They must each consist of a single molecule that has no covalent bonds to the receptor,
with no accompanying fragments, such as counter ions and solvent molecules.
4. They must have all their hydrogens (filled valences).
5. They must have an appropriate protonation state for physiological pH values (around 7).
For example, carboxylic acids should be deprotonated and aliphatic amines should be
protonated. Otherwise a neutral aliphatic amine could improperly act as a hydrogen-bond
acceptor in the docking calculations, or could occupy a hydrophobic region without
incurring the large desolvation penalty that XP Glide docking would have assessed if the
amine had been properly protonated.
Protonation states are particularly crucial when the receptor site is a metalloprotein such
as thermolysin or a MMP. If the metal center and its directly coordinated protein residues
have a net charge, Glide assigns a special stability to ligands in which anions or polar
groups coordinate to the metal center.
6. They must be supplied in Maestro, SD, Mol2, or PDB format.
Glide 6.7 User Manual
21
Chapter 3: Protein and Ligand Preparation
Maestro transparently converts SD, MacroModel, Mol2, PDB, and other formats to Maestro format during structure import. However, Glide has no direct support for other
formats, so you should ensure that your structures are in Maestro, SD, Mol2, or PDB format before starting Glide jobs.
All of the above conditions can be met by using LigPrep to prepare the structures. Use of
LigPrep is described in the next section.
3.4.1
Using LigPrep for Ligand Preparation
The Schrödinger ligand preparation product LigPrep is designed to prepare high quality, allatom 3D structures for large numbers of drug-like molecules, starting with 2D or 3D structures
in SD, Maestro, or SMILES format. LigPrep can be run from Maestro or from the command
line. For detailed information on LigPrep, see the LigPrep User Manual.
To run LigPrep, you must have a LigPrep license. The MacroModel commands premin and
bmin require LigPrep licenses when run in a LigPrep context, and are limited to a restricted set
of commands when run using a LigPrep license. LigPrep can be run from Maestro or from the
command line.
The LigPrep process consists of a series of steps that perform conversions, apply corrections to
the structures, generate variations on the structures, eliminate unwanted structures, and optimize the structures. Many of the steps are optional, and are controlled by selecting options in
the LigPrep panel or specifying command-line options. The steps are listed below.
1. Convert structure format.
2. Select structures.
3. Add hydrogen atoms.
4. Remove unwanted molecules.
5. Neutralize charged groups.
6. Generate ionization states.
7. Generate tautomers.
8. Filter structures.
9. Generate alternative chiralities.
10. Generate low-energy ring conformations.
11. Remove problematic structures.
12. Optimize the geometries.
22
Schrödinger Software Release 2015-2
Chapter 3: Protein and Ligand Preparation
13. Convert output file.
The LigPrep panel allows you to set up LigPrep jobs in Maestro. Choose LigPrep from the
Applications menu to open the panel. For details of panel options and operation, see Chapter 2
of the LigPrep User Manual.
The simplest use of LigPrep produces a single low-energy 3D structure with correct chiralities
for each successfully processed input structure. LigPrep can also produce a number of structures from each input structure with various ionization states, tautomers, stereochemistries, and
ring conformations, and eliminate molecules using various criteria including molecular weight
or specified numbers and types of functional groups present.
The default options in the LigPrep panel remove unwanted molecules, add hydrogens, and
minimize the ligand structure (performing a 2D-3D conversion, if necessary). Below are notes
on panel options that produce more than one output structure per input structure.
The Ionization options allow you to generate all the ligand protonation states that would be
found in the specified pH range. The Ionization options are:
• Retain original state—Retain the original ionization state rather than attempting to ionize
the ligand.
• Neutralize—Neutralize the ligand by adding or removing protons from ionizable groups.
• Generate possible states at target pH target +/- range—This is the default, and can generate several different output structures for each input structure. The default pH target is 7.0
with a +/- range of 2.0, so the default pH range is 5.0 – 9.0. Both the target and range settings can be changed. You can use either the ionizer or Epik to generate ionization
states. Epik generally yields more accurate states, because it uses a more sophisticated
algorithm and performs ionization and tautomerization together. In addition, Epik generates a set of states that are more appropriate for metal binding when you select Add metal
binding states, and calculates penalties for these states to use in the docking score. Epik is
a separate product, so you must purchase this product to use it.
Generate low energy ring conformations: number per ligand. The default is to generate only the
lowest energy conformation.
Desalt is selected by default.
Generate tautomers is selected by default. The tautomerizer generates up to 8 tautomers per
ligand, selecting the most likely tautomers if more than 8 are possible. If you are sure that the
input structures are already in the correct tautomeric form for docking to a particular target,
then the tautomerizer should be turned off by deselecting Generate tautomers.
Glide 6.7 User Manual
23
Chapter 3: Protein and Ligand Preparation
The stereoizer can generate two stereoisomers per chiral center in the ligand, up to a specified maximum. There are three Stereoisomers options:
The first two options, Retain specified chiralities (the default) and Determine chiralities
from 3D structure, generate both isomers only at chiral centers where chirality is unspecified or indeterminate; centers with known chirality retain that chirality. The difference is
that Retain specified chiralities takes its chirality data from the input file (SD or Maestro),
while Determine chiralities from 3D structure ignores input file chiralities and takes chirality information from the 3D geometry.
Generate all combinations varies the stereochemistry up to a maximum number of structures specified by Generate at most max per ligand. The default maximum is 32.
3.4.2
Using Other Programs for Ligand Preparation
If you prefer to prepare the ligands with other programs, you can do so. Schrödinger software
installations include a number of utilities that can be used to perform some of the above tasks.
These utilities are also used by LigPrep. One of these, the Ionizer, can be used to prepare
ligands in the required protonation states. Some of the other tasks can be performed as follows:
• Hydrogen atoms can be added in Maestro with either the Add H toolbar button:
or the Add Hydrogens - Advanced panel (choose Edit > Hydrogen Treatment > Advanced).
Hydrogen atoms can also be added (or removed) using the utility applyhtreat, which is
described in Section 4.1 of the General Utilities manual.
• Structure file format conversion can be done from the command line with utilities such as
structconvert, pdbconvert, and sdconvert—see Section 1 of the General Utilities
manual.
24
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 4
Chapter 4:
Receptor Grid Generation
Glide searches for favorable interactions between one or more ligand molecules and a receptor
molecule, usually a protein. The shape and properties of the receptor are represented on a grid
by several different sets of fields that provide progressively more accurate scoring of the ligand
poses. For receptors that adopt more than one conformation on binding, you might want to
prepare grids for each conformation, to ensure that possible actives are not missed. Glide can,
however, handle different hydroxyl conformations with a single grid generation.
The receptor grid can be set up and generated from the Receptor Grid Generation panel. The
options in each tab of this panel allow you to define the receptor structure by excluding any
cocrystallized ligand that may be present, determine the position and size of the active site as it
will be represented by receptor grids, set up Glide constraints, and set up flexible hydroxyl
groups. Ligand docking jobs cannot be performed until the receptor grids have been generated.
Receptor grid generation requires a “prepared” structure: an all-atom structure with appropriate bond orders and formal charges. Information on structure preparation is given in
Chapter 3.
The force field used for grid generation is the OPLS_2005 force field, which allows a proper
treatment of metals and has a wider range of atom types defined than its predecessor,
OPLS_2001.
A tutorial example of setting positional and H-bond constraints is given in Chapter 2 of the
Glide Quick Start Guide.
4.1
The Receptor Grid Generation Panel
To open the Receptor Grid Generation panel, choose Receptor Grid Generation from the Glide
submenu of the Applications menu. The Receptor Grid Generation panel has five tabs, which
you use to specify settings for the receptor grid generation job:
•
•
•
•
•
Receptor
Site
Constraints
Rotatable Groups
Excluded Volumes
These tabs are described in later sections of this chapter.
Glide 6.7 User Manual
25
Chapter 4: Receptor Grid Generation
Below the tabs are controls for starting the job, which are described in general in Section 1.2
on page 3. Some details are given here.
The Run button starts the job immediately with the current settings for the job, including the
job name and the host.
The Job name text box allows you to specify the job name without opening the Job Settings
dialog box. When you click the Start button, the job is run with the name in this text box.
The Settings button opens the Receptor Grid Generation - Job Settings dialog box.
Figure 4.1. The Job Settings dialog box for grid generation jobs.
The dialog box has two sections: Output and Job. In the Output section you can specify the
location for the grid files, which are written as zip files.
In the Job section, you can name your job by typing the name into the Name text box, and you
can specify the Host to run the job. To save settings, click OK; to save settings and start the job,
click Run. For more information on this dialog box, see Section 2.2 of the Job Control Guide.
The Settings button has an option menu (arrow to the right of the icon) with the following
items:
Write
Choose Write to write the input files without starting the job. The job name is taken from the
Job name text box. This name is used to create a directory in which the input files are written,
and to name the files. The Glide input file is written to the file jobname.in, and the receptor is
written to the file jobname.maegz. A script is also written to the file jobname.sh, containing
the command for running the job in a Unix shell.
Reset Panel
Choose Reset Panel to restore the default settings in all tabs.
26
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
4.2
The Receptor Tab
In this tab you define the part of the Workspace system for which receptor grids should be
calculated. You can also scale receptor atom van der Waals radii in this tab, and choose
whether to use partial charges from the force field or from the input structure. The tab has three
sections, Define receptor, Van der Waals radii scaling, and an unlabeled section with several
options in it. Some more advanced settings that are not commonly needed are in a dialog box,
which you open by clicking Advanced Settings.
Figure 4.2. The Receptor tab of the Receptor Grid Generation panel.
4.2.1
Defining the Receptor
The Define receptor section contains options for defining the part of the system in the Workspace to be treated as the receptor. If only the receptor is included in the Workspace, and no
ligand is present, you can ignore the Define receptor options.
• If the Workspace includes both a receptor and a ligand, use these options to pick the
ligand molecule. The ligand will be excluded from receptor grid generation. Everything
not defined as the ligand will be treated as part of the receptor. The ligand can be identified either as a molecule or as an entry in the Workspace.
• If you want to use a binding site from a SiteMap calculation, you can do so by treating it
as a ligand and including it in the Workspace. Each site from SiteMap is a separate entry.
Glide 6.7 User Manual
27
Chapter 4: Receptor Grid Generation
However, the site points in the site are not connected, so the site must be selected as an
entry.
To select the ligand, ensure that Pick to identify ligand molecule is selected, choose an option
from the option menu, then pick an atom in the ligand molecule or SiteMap site. The ligand (or
site) is now distinguished from the receptor. If Show markers is selected, the ligand molecule is
marked with green markers. Deselect the option to remove the markers.
4.2.2
Global Van der Waals Radius Scaling
Glide does not allow for receptor flexibility in docking (apart from hydroxyl rotations), but
scaling of van der Waals radii of nonpolar atoms, which decreases penalties for close contacts,
can be used to model a slight “give” in the receptor and the ligand. Scaling of other interactions
can also help to model flexibility of parts of the receptor.
If you have a receptor in which there is substantial movement upon docking, such as a change
in side-chain conformation, backbone location or loop conformation, you should consider
docking to multiple protein conformations. You can also use the Induced Fit Docking protocol
to account for receptor flexibility. This protocol uses Glide and Prime, and is much more
computationally demanding than Glide docking alone. It is therefore mainly useful for docking
a small number of ligands. See the document Induced Fit Docking for more information.
If you have a receptor that has Ser, Thr, or Tyr residues in the active site and the rotation of the
hydroxyls on these groups is important, you can specify the hydroxyls as rotatable groups in
the Rotatable Groups tab—see Section 4.5 on page 41 for details.
Glide has two means of accounting for protein flexibility by scaling of van der Waals radii. The
first is described here; the second is described in the next section.
For nonpolar receptor atoms, you can use the controls in the Van der Waals radius scaling
section to scale the van der Waals radius of those receptor atoms. The definition of nonpolar
atoms is determined by a partial charge threshold that you can set. For ordinary Glide docking,
it is recommended that receptor radii be left unchanged, and any scaling be carried out on
ligand atoms. Receptor scaling is probably most useful when the active site is tight and encapsulated.
The Scale by text box specifies the scaling factor. Van der Waals radii of nonpolar receptor
atoms are multiplied by this value. The default value is 1.00, for which no scaling is done.
Scaling of van der Waals radii is performed only on nonpolar atoms, defined as those for which
the absolute value of the partial atomic charge is less than or equal to the number in the text
box. Since this is an absolute value, the number entered must be positive. The default is 0.25.
28
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
4.2.3
Selection of Partial Charges
You can use partial charges from the input structure, rather than from the force field, by
selecting Use input partial charges. This option is useful if, for example, you have obtained
improved partial charges around the active site, such as those from a QSite calculation or a
QM-Polarized Ligand Docking calculation.
4.2.4
Grids for Peptide Docking
If you want to dock peptides with Glide, the grid must be prepared for this purpose. To do so,
select Generate grid suitable for peptide docking. The grid can only be used in SP-Peptide
docking mode.
4.2.5
Per-Atom Van der Waals Radius and Charge Scaling
For a more flexible method of softening the receptor potential, you can specify van der Waals
radii and charges on a per-atom basis. This is done in the Per-atom van der Waals radius and
charge scaling section of the Receptor - Advanced Settings dialog box, which you open by
clicking Advanced Settings in the Receptor tab. If you specify per-atom scaling, the global
scaling of nonpolar atoms is only performed on those atoms for which per-atom scaling factors
are not specified.
Figure 4.3. The Receptor - Advanced Settings dialog box.
Glide 6.7 User Manual
29
Chapter 4: Receptor Grid Generation
There are three choices for the source of per-atom scale factors:
• None—do not use per-atom scale factors.
• Read from input structure file—Read the radius and charge scaling factors from the input
file for the receptor. This file must be in Maestro format, and must have the properties
VdW radius scale factor (r_glide_atom_vdwscale) and Charge scale factor
(r_glide_atom_coulscale) defined. These properties must have a value for each
atom in the receptor. They are defined if you have previously set per-atom scaling factors.
• Specify for selected atoms—Specify the scaling factors for selected atoms. You can make
multiple selections and apply scaling factors to each, using the Select atoms for scaling
factor picking tools. The selections should be mutually exclusive. A row is added to the
table for each pick, with the scale factors specified in the van der Waals radius scale factor text box and Charge scale factor text box. You can subsequently edit the factors in the
table.
4.2.6
Hydrogen Bonds with Aromatic Hydrogens and Halogens
In addition to normal hydrogen bonds, you might want to account for other noncovalent interactions that are generally weaker, but could be important. You can choose to include three
types of interactions in the grid generation: hydrogen bonds to aromatic hydrogens, hydrogen
bonds to halogens (halogen as an acceptor), and halogen bonds (halogen as a donor). The
options to include these interactions are in the Count aromatic H and halogen H-bonds section
of the Receptor - Advanced Settings dialog box, which you open by clicking Advanced
Settings in the Receptor tab. (Strictly speaking, halogen bonds aren’t H-bonds, but they are
included here as a class of weak noncovalent interactions.) Currently you can only choose to
have halogens treated as donors or as acceptors, but not both.
4.3
The Site Tab
The settings in the Site tab determine where the scoring grids are positioned and how they are
prepared from the structure in the Workspace. To make these settings, you need to understand
how Glide sets up grids.
Glide uses two “boxes” to organize the calculation:
• The grids themselves are calculated in the space defined by the enclosing box, or grid box
or outer box. This is also the box within which all the ligand atoms must be contained.
• During the site point search, the ligand center is allowed to move within the ligand diameter midpoint box, or inner box. This box gives a truer measure of the effective size of the
search space. However, ligands can move outside this box during grid minimization.
30
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
The only requirement on the grid box is that it is big enough to contain all ligand atoms when
the ligand center is placed at an edge or vertex of the inner box. Grid boxes that are larger than
this are not useful: they take up more space on disk and in memory for the scoring grids, which
take longer to compute. The maximum size of the grid box is 80 Å.
The ligand center is defined in a rigid-docking run as the midpoint of the line drawn between
the two most widely separated atoms. The definition changes slightly for flexible docking,
where the ligand center becomes the midpoint between the two most widely separated atoms of
the core region—the part of the ligand remaining after each of the end-groups has been
stripped off at the terminal end of the connecting rotatable bond.
The two boxes share a common center. Thus, the operations in the tab that center one box also
center the other. Information on the two boxes is written to the Maestro file for the receptor.
Each rigidly docked ligand or flexibly docked conformation has an associated length, L, which
can be defined as twice the distance from the ligand center to its farthest atom. The required
relationship between L and the lengths E and B of the outer and inner boxes for successful
placement of the ligand center anywhere within the inner box is:
E≥B+L
The grid box must be large enough in each dimension to hold the length of the inner box plus
the maximum length of any ligand. If a larger ligand is encountered, not all positions for the
center of the ligand in the inner box are accessible. The effective inner box for that ligand will
be smaller than the dimension nominally specified. In any docking job using these receptor
grids, ligands are confined to the grid box.
If the structure in the Workspace consists of a receptor and the ligand molecule you identified
in the Receptor tab, Glide uses the position and size of the ligand to calculate a default center
and a default size for the grid box. When you open the Site tab, the Workspace displays the
center of the grid box as a set of coordinate axes colored bright green, and the boundaries of
the region as a purple wire-frame cube.
If the Workspace includes a SiteMap binding site, you might want to reduce the size of the grid
box, because it is likely that the site is larger than defined by a ligand. You might also want to
specify the center of the box by selection of a few residues from the receptor, as the centroid of
the site might not be in the optimal location. This is particularly so if the site is not well defined
or the site points extend over a broad region.
If the Workspace structure consists of a receptor only, there is no default center for the grid
box. The box will not be displayed until you have specified a grid center by selecting residues.
By default, the purple grid box outline and the green axes at the center are displayed when you
enter the tab. Deselect Display Box to undisplay the box and its center.
Glide 6.7 User Manual
31
Chapter 4: Receptor Grid Generation
Figure 4.4. The Site tab of the Receptor Grid Generation panel.
4.3.1
Selecting a Box Center
Select one of the options under Center to determine how the center of the grid is defined:
• Centroid of Workspace ligand
This option centers grids at the centroid of the ligand molecule that was defined in the
Receptor tab, also called the Workspace ligand. If a Workspace ligand has been defined,
this option is the default. The Advanced Settings button is available with this option.
• Centroid of selected residues
This option centers grids at the centroid of a set of residues that you select. With this
option you can define the active site (where grids should be centered) with only the receptor in the Workspace. The Specify Residue button is only available when you choose this
option; the Advanced Settings button is not available with this option.
To select the residues, click Specify Residues. The Active Site Residues dialog box
opens. Using the picking controls, you can pick the residues that best define the active
site. The list of selected residues is displayed in the table. You can delete residues by
selecting them in the list, and clicking Delete. To delete all residues, click Delete All.
32
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
Figure 4.5. The Active Site Residues dialog box.
• Supplied X, Y, Z coordinates
This option centers the grid at the Cartesian coordinates that you specify in the X, Y, and Z
text boxes. These text boxes are only available when you choose this option. The
Advanced Settings button is available with this option.
4.3.2
Setting the Box Sizes
The Size section provides options for the size of the grid box. The default option is Dock
ligands similar in size to the Workspace ligand when a ligand is used for the grid center. This
choice is suitable when the ligands to be docked are of the same size as, or smaller than, the
Workspace ligand. If you expect to dock larger ligands, or if there is no Workspace ligand,
select Dock ligands with length <= and use the slider to choose an appropriate maximum ligand
length. The slider is set to 20 Å by default. If you use the centroid of specified residues, or
supplied coordinates, Dock ligands with length <= is the default option.
To change the size of the inner box, click Advanced Settings. The Site - Advanced Settings
dialog box opens, and the inner box, or ligand diameter midpoint box, is displayed as a cube
outlined in bright green. The diameter midpoint of each docked ligand remains within this box
in the site-point search stage, but can move outside this box in the grid minimization stage (see
Figure 2.2 on page 9). You can use the Size sliders to increase or decrease the dimensions of
each side of the box. The default is 10 Å on each side; the allowed range is 6 Å to 14 Å.
A larger inner box can be useful to allow ligands to find unusual or asymmetric binding modes
in the active site. Conversely, if the default inner box allows ligands to stray into regions you
know to be unfruitful, you can confine their midpoints to a smaller box, eliminating some of
the less useful poses and saving calculation time. Changing the shape of the box can be useful
when the active site is spatially extended in one or more directions.
Glide 6.7 User Manual
33
Chapter 4: Receptor Grid Generation
Figure 4.6. The Site - Advanced Settings dialog box.
4.4
The Constraints Tab
The Constraints tab of the Receptor Grid Generation panel is used to define Glide constraints
for the receptor grids to be generated. Glide constraints are receptor-ligand interactions that
you believe to be important to the binding mode, based on structural or biochemical data.
Setting constraints enables Glide to screen out ligands, conformations, or poses that do not
meet these criteria early on in their evaluation for docking suitability.
There are five types of Glide constraints that are set up during grid generation: positional
constraints, NOE constraints, H-bond constraints, metal constraints, and metal coordination
constraints. Two other types, core constraints and torsional constraints, only depend on the
ligand and are specified during docking setup (see Section 5.5 on page 58 and Section 5.7 on
page 69).
• A positional constraint is a requirement that one or more ligand atoms occupy a spherical
volume that is centered at a particular position. This is a maximum distance constraint.
• A NOE (nuclear Overhauser effect) constraint is a requirement that one or more ligand
atoms occupy a spherical shell (the region between two spheres) that is centered at a particular position. This is a constraint to a distance range.
• An H-bond constraint is a requirement that a particular receptor-ligand hydrogen bond be
formed.
• A metal constraint is a requirement that a particular metal-ligand interaction is present
when the ligand is docked. The ligand atom must lie in a sphere around a specified receptor metal atom, and therefore the constraint on the ligand atom has no directionality.
• A metal coordination constraint is a requirement that a ligand atom lie within a given distance of an optimal coordination site for a metal atom. It differs from a metal constraint in
that the constraint sphere is centered on a potential ligand coordinating atom rather than
on the metal, and is thus directional.
34
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
The constraints tab has three subtabs, Positional/NOE, H-bond/Metal, and Metal Coordination,
for setting up the constraints. The number of constraints of each type is reported on the subtab
tab, and the total number of constraints is reported at the top of the tab. The maximum number
of constraints that you can define for a given grid is ten constraints, distributed among positional, H-bond, metal, and metal coordination constraints.
When constraints setup is complete and the grid generation job is run, Glide writes a file
containing the information about the constraints. Subsequent docking jobs use this file to determine whether a given ligand pose satisfies the constraints. If the base name for writing grid
files is gridbase, the constraints file is named gridbase.cons. Not all of these constraints are
used in a given docking job: when you set up the docking job, you can choose which
constraints to apply.
To use Glide constraints, you must define the constraints during grid generation setup and then
apply the constraints during docking setup (see Section 5.6 on page 61). Any Glide constraints
that you want to apply in docking must be defined when the receptor grids are generated.
Note:
You cannot use flexible receptor groups for constraints.
When you are setting up constraints, it may be helpful to undisplay most of the receptor,
leaving only residues within a short distance of the ligand visible.
Figure 4.7. The Constraints tab of the Receptor Grid Generation panel showing the
Positional/NOE subtab.
Glide 6.7 User Manual
35
Chapter 4: Receptor Grid Generation
4.4.1
Setting Positional and NOE Constraints
Positional constraints define a region that must contain a particular kind of ligand atom. The
specific kind of atom is defined during docking setup, using SMARTS patterns. Positional
constraints allow you to require interactions between any kind of receptor and ligand atoms,
while at the same time placing tighter restrictions on the ligand atom position than is typical
with other constraint types.
For example, a hydrogen-bond acceptor in the receptor might be capable of forming hydrogen
bonds in two directions, but only one of these results in good binding. While setting an H-bond
constraint in this case allows a ligand hydrogen atom to lie in either of these directions, a positional constraint can require it to be in the “good” direction. The constraint could be set by
selecting a hydrogen atom that bonds in the “good” direction to define the position of the
constraint. For reasons such as this, we suggest that you display a model ligand in the Workspace to aid in selecting appropriate positions for constraints.
NOE constraints are similar to positional constraints, but require that the ligand atoms lie in a
given distance range from the constraint center, i.e. in the “shell” between two spheres.
To add a positional or NOE constraint, click New. This button opens the New Position/NOE
dialog box, in which you can pick atoms with the standard picking controls to define a
constraint center, name the constraint, and specify its radius or minimum and maximum
distances. To choose the constraint type, select Position or NOE under Constraint type.
Figure 4.8. The New Position/NOE dialog box.
The center of the sphere or spheres for a constraint is the centroid of the selected atoms, and
must be inside the grid box. While picking is in progress, the constraint is marked with a gray
sphere. If you make the constraint an NOE constraint, both spheres are displayed. When you
click OK, the constraint is added to the Positions table if it is inside the grid box; otherwise a
warning is displayed.
36
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
To delete a constraint, select it in the Positions table and click Delete; to delete all constraints,
click Delete all.
The Positions table displays the constraints you have chosen, giving the name, coordinates, and
radii of the constraint spheres for each constraint. The coordinates and the radii are given in
angstroms. You can select a single constraint in the table, and edit the name, coordinates, and
sphere radii by clicking in the table cell and changing the value. If you want to convert a positional constraint to an NOE constraint, you can do so by providing a minimum distance.
To view the constraints in the Workspace, select Markers. The selected constraint is marked by
one or two yellow spheres. The other positional or NOE constraints are marked by red spheres.
If Markers is selected, selecting the Label positions option displays the name of the constraint
in the Workspace. The labels are colored the same as the constraints.
4.4.2
Setting H-Bond and Metal Constraints
Up to ten symmetry-distinct receptor atoms can be chosen as possible H-bond or metal
constraint sites.
For hydrogen-bonding interactions, the receptor atom must be a polar hydrogen (including
thiol H in cysteine), nitrogen, or oxygen. If you choose an atom with one or more symmetryequivalent atoms in its functional group, the symmetry-equivalent atoms are all selected as
well by default, and collectively count as one constraint. For example, if you create a constraint
by picking one oxygen atom of a carboxylate group, Glide includes the other oxygen atom in
the same constraint. A ligand interaction with either oxygen atom satisfies that single
constraint. However, you can turn off the use of symmetry so that only the chosen atom is used.
For metal-ligand interaction constraints, the receptor atom must be a metal ion.
The receptor atoms selected must also be close enough to the ligand that satisfying the
constraints is possible. You do not need to specify limits on distances or angles between
receptor and ligand atoms for the constraint: Glide sets these values internally, to H-acceptor
distances of 1.2 to 2.5 Å; donor angles greater than 90°, and acceptor angles greater than 60°.
For metal constraints the ligand-metal distance must be no greater than the sum of the van der
Waals radii of the metal and ligand atoms plus 0.4 Å.
These values are looser than those employed by Maestro. General distance requirements are
incorporated using the grid box for the ligand. The receptor atoms selected for constraints must
be inside the grid box (which is displayed in purple) or within bonding range of it.
To display hydrogen bonds in the Workspace, choose Inter from the Display H-bonds button
menu and click on a ligand atom. The hydrogen bonds between the ligand and the receptor are
displayed. This should make it easier to locate the relevant receptor atoms. If you want to use
Glide 6.7 User Manual
37
Chapter 4: Receptor Grid Generation
the looser criteria for hydrogen bonds employed by Glide, you can do so in the Preferences
panel. Choose Maestro → Preferences and under Nonbonded Interactions – Criteria, click
Glide in the H-Bonds section.
Figure 4.9. The Constraints tab of the Receptor Grid Generation panel showing the
H-bond/Metal subtab.
To set H-bond or metal constraints, ensure that Pick atoms is selected in the H-bond/Metal
constraints tab, and pick the desired atoms in the Workspace. By default, symmetry-equivalent
atoms such as the other oxygen in a carboxylate group are included in the constraint. If you
only want to use the picked atom for the constraint, clear the check box in the Use Symmetry
column. If Show markers is selected, a red cross and red padlock appear next to each atom
picked. If the picked atom is one of a set of symmetry-equivalent atoms, all the atoms in the set
are marked, provided Use Symmetry is selected. If Show markers is selected, selecting Label
atoms displays the constraint name in the Workspace.
As you select atoms in the receptor, they appear in the Receptor atoms table. The Atom column
gives detailed information on the atom selected in the format
atom-number:chain:residue-name residue-number :atom-name : symmetry-set
38
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
where
•
•
•
•
•
•
atom-number is the Maestro atom number
chain is the chain name
residue-name is the name of the residue
residue-number is the residue number and the insertion code, if any
atom-name is the PDB atom name
symmetry-set is the atom name or symmetry-equivalent atom set
for example,
341:C:ASN 239 : OD1 : ODn
If the picked atom is part of a symmetry-equivalent set and Use Symmetry is selected, its identification is followed by square brackets enclosing the number and name of each atom in the
set, separated by commas.
The Name column gives a label that is used when displaying the possible constraints from the
Ligand Docking panel. The format of the name is:
chain:residue-name:residue-number:atom-name(type)
where the quantities except type are defined above, and type is either hbond or metal, for
example,
C:ASN:239:OD1(hbond)
This name can be changed by editing the table cell.
The Use Symmetry column indicates whether symmetry-equivalent atoms are included with
the picked atom for a constraint or not.
To delete a single H-bond or metal constraint, select it in the list and click Delete; to delete all
the listed constraints, click Delete All.
4.4.3
Setting Metal Coordination Constraints
Metal coordination constraints require a ligand atom to lie within a specified distance of a
coordination site, which is the location that a ligand atom should occupy for optimal bonding
with the metal. For each metal, the possible coordination sites are identified, and a constraint
sphere is placed at the ideal location of a ligand donor atom at the available site. This differs
from the metal constraint feature, in which a constraint sphere is placed on the metal. The
metal coordination constraint has directionality, whereas the metal constraint has none. You
can choose to use any or none of the sites found for a given metal.
Glide 6.7 User Manual
39
Chapter 4: Receptor Grid Generation
Figure 4.10. The Constraints tab of the Receptor Grid Generation panel showing the
Metal Coordination subtab.
To choose a metal for a coordination constraint, select Pick metal atoms (the default), then pick
a receptor metal atom in the Workspace. (The atom is identified in the status bar when the
pointer is over the atom.) When a metal is picked, the possible coordination sites are identified
and rows are added to the Receptor metal atoms table with the coordinates of the site and the
maximum distance. Each of the possible coordination constraints is shown in the Workspace as
a sphere centered at the coordination site, with the radius equal to the maximum distance. If
there is more than one possible coordination geometry, all geometries are listed.
All of the sites for the first coordination geometry are selected for use by default, but you can
choose which ones you want to use by checking or clearing the check box in the Use column.
When you select a row in the table, all the sites associated with the metal atom for that coordination geometry are also selected, and highlighted in the Workspace in red. Any other spheres
are colored gray. You can edit the constraint name and the coordinates and radii of the
constraint spheres by clicking in the appropriate table cell and changing the value, or you can
delete the set of sites for a given metal by clicking Delete. You cannot delete individual sites:
instead, you should clear the Use check box to remove them from consideration. When the
check box is cleared, the sphere in the Workspace turns gray.
It may be necessary to adjust the orientations of the sites from their initial locations, so that the
constraint spheres cover the desired regions of space. As coordination orientations can vary
40
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
considerably in a complex structure such as a protein, it may be necessary to make some
adjustments. To do this, select Rotate, and then rotate the orientation of the coordination sites
around the metal with the usual actions for Workspace rotation.
You can hide the constraint spheres in the workspace by deselecting Show markers. You can
also hide the label on the constraint spheres by deselecting Label regions. This label is the
constraint name, and is the same for all constraint spheres belonging to a particular metal atom.
4.5
The Rotatable Groups Tab
The hydroxyl groups in residues such as Ser, Thr, and Tyr and the thiol group in Cys can adopt
different orientations with different ligands. Glide can allow such groups to adopt different
orientations when ligands are docked, to produce the most favorable interaction. For Ser and
Thr, the hydroxyls can be oriented in any of the three local minima, and likewise for the thiol
of Cys; for Tyr, they can be in either of the two local minima. The Rotatable Groups tab allows
you to choose which of these groups should be treated flexibly.
Figure 4.11. The Rotatable Groups tab of the Receptor Grid Generation panel.
The Available groups table is loaded by default with a list of all rotatable groups within the grid
box. The group is identified by the hydrogen atom number and the residue (chain, number,
name). By default, none of them are selected for use. You can then choose the groups that you
want to include in the grid generation either by selecting them in the Allow rotation column of
Glide 6.7 User Manual
41
Chapter 4: Receptor Grid Generation
the table, or by selecting Pick groups and picking the hydrogens of the groups in the Workspace. The selected groups are marked in the Workspace. Clicking a second time, either in the
column or in the Workspace, deselects the group. Clicking Deselect All deselects all groups
(which has the effect of turning off the use of rotatable groups).
If you have previously prepared grids with flexible groups, and you are using the receptor from
a file that was written by Glide, the existing selection is loaded by default. The rotatable group
information is added to the receptor structure when it is written out by Glide.
You cannot create constraints to flexible groups.
When docking ligands with flexible groups, the time taken is longer than for a non-flexible
docking calculation, but significantly shorter than docking with all possible combinations of
flexible group orientations. For example, with 4 flexible groups, an SP docking run takes about
twice as long as a run with no flexible groups, and an XP docking run takes about 4 times
longer. If each of these groups can take two possible orientations, there are 16 combinations.
To dock to all of these combinations without allowing flexibility would therefore require 16
grid calculations and 16 non-flexible docking runs, and the results from the individual runs
would have to be collated at the end.
Note:
4.6
Once you have set up a grid with flexible groups, the flexibility is used in docking, and
cannot be turned off.
The Excluded Volumes Tab
In some situations, you might want to prevent ligands from occupying certain regions of space.
For example, if you have a pocket near the active site where ligands are known not to bind, you
might want to stop ligands from occupying that pocket. Another situation is searching for
ligands that might be immune to drug-resistant mutations, to check the alignment by ensuring
that the drug only occupies the space that is occupied by the substrate. A third case is where
parts of a protein are missing, and you want to prevent the ligand from occupying that region.
In the Excluded Volumes tab, you can set up regions of space from which the ligands will be
excluded during docking. These regions are composed of a set of spheres, which you can
define by picking atoms for the centroid of a sphere (e.g. atoms on either side of a pocket you
want to exclude), and then setting the radius of the sphere. You would not normally need to
place spheres on regions occupied by the protein, because these are already excluded in the
docking process (due to the potentials).
42
Schrödinger Software Release 2015-2
Chapter 4: Receptor Grid Generation
Figure 4.12. The Excluded Volumes tab of the Receptor Grid Generation panel.
To set up an excluded volume:
1. Click New.
The New Excluded Volume dialog box opens.
Figure 4.13. The New Excluded Volume dialog box.
2. Pick atoms in the Workspace to define the center of the excluded volume sphere.
When you start to pick, a sphere is placed in the Workspace. The sphere center is placed
at the centroid of the picked atoms, and moves to the new centroid as you pick more
atoms.
3. Enter a name for the excluded volume in the Name text box.
Glide 6.7 User Manual
43
Chapter 4: Receptor Grid Generation
4. Enter the sphere radius in the Radius text box.
5. Click OK.
The dialog box closes, and a new row is added to the Excluded volumes table.
You can change the sphere parameters by editing the table cells for the excluded volume. You
can define as many excluded volumes as you like.
Phase also allows you to define excluded volumes for a pharmacophore hypothesis. You can
import a Phase excluded volume file (.xvol) to use with a Glide grid, by clicking Import and
navigating to the file. Importing Phase excluded volumes is only likely to be useful if the
ligand you are using as the reference ligand for Glide is the same as that used for the Phase
hypothesis, or at least has pharmacophore features that superimpose well on the Glide ligand.
To check the alignment, you should include both the Phase hypothesis with its reference ligand
and the Glide native ligand in the Workspace.
Glide excluded volume files (.gxvol) cannot be used with Phase, however, because the format
is not compatible, and they contain information other than the sphere coordinates and radii.
44
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 5
Chapter 5:
Ligand Docking
Glide ligand docking jobs require a set of previously calculated receptor grids and one or more
ligand structures. The ligand structures must satisfy the conditions listed on page 21. Information on setting up grid generation jobs is given in Chapter 4. For docking tutorial examples, see
Chapter 3 of the Glide Quick Start Guide.
Preparation of the ligands before docking is strongly recommended. LigPrep or MacroModel
can be used to prepare ligands—see Chapter 3 for more information.
If a correct Lewis structure cannot be generated for a ligand, it is skipped by the docking job.
Glide also automatically skips ligands containing unparametrized elements, such as tin, or
atom types not supported by the OPLS force fields, such as explicit lone pair “atoms.”
This chapter contains a detailed description of the Ligand Docking panel in Maestro and each of
its tabs, including instructions for applying Glide constraints, using extra-precision Glide
docking (Glide XP), and distributed processing.
5.1
The Ligand Docking Panel
To open the Ligand Docking panel, choose Ligand Docking from the Glide submenu of the
Applications menu. The Ligand Docking panel has several tabs:
•
•
•
•
•
•
Ligands
Settings
Core
Constraints
Torsional Constraints
Output
These tabs are described in the following sections of this chapter. You use the options in these
tabs to specify settings for the ligand docking job. When you have completed your setup, the
Job toolbar at the bottom of the panel allows you to process your job.
• Specify the name for the job in the Job name
• Click the Settings button (gear icon) to make other job settings.
• Click the Settings button arrow and choose Write to write the input files to disk without
starting the job, or choose Reset to discard settings and restore the defaults in all tabs.
Glide 6.7 User Manual
45
Chapter 5: Ligand Docking
When you write the input files, you can choose a standard job name, type in a name, or browse
to the desired location. The input files are written to the current working directory by default.
The Ligand Docking - Job Settings dialog box provides controls for the running of the job. You
can name the job by entering text in the Name text box. If you want to incorporate the results
into the current Maestro project, you can select an option from the Incorporate option menu.
Incorporation can take some time if you have a large number of poses.
Figure 5.1. The Job Settings dialog box for ligand docking jobs.
You can also specify options for distributed processing, by splitting the job into subjobs and
distributing the subjobs over multiple processors. To separate the docking job into a specific
number of subjobs, select Exactly under Separate job into and enter the number of subjobs in
the text box. To separate the docking job into subjobs of a maximum size (number of ligand),
select Subjobs with no more than and enter the maximum number of ligands in the text box. To
use multiple processors, select the hosts and enter number of processors to use on each host in
the host table. You can edit the cells in the Use column to set the number of processors to use
for each host. For details on the use of this table, see Section 2.2 of the Job Control Guide.
The number of subjobs cannot be set to less than the number of processors. For optimal load
balancing and for restartability, the number of subjobs should be several times the number of
processors—even if you are running on a single processor. It is recommended that you split
any docking job into subjobs with no more than about 50,000 ligands, to minimize the loss of
work in the event of a job failure (which could be due to network problems or hardware
failure). Glide tries to restart failed subjobs, and if there is only one subjob, it restarts the failed
job from the beginning.
46
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
See Section 2.2 of the Job Control Guide for more information on the Job toolbar and the Job
Settings dialog box.
5.2
Specifying the Receptor Grid
To specify the receptor grid for the docking job, click Browse in the Receptor grid section of
the Settings tab to open a file selector and choose a compressed grid archive (.zip) or a grid
file (.grd). Note that the grid archive cannot be renamed, otherwise the docking job will fail.
The file name is displayed in the Receptor grid text box. You can also enter the file name
directly into the text box. The grid file must be accessible on the local host; if it is not, the job
will not be submitted. If you want to use a grid that is not accessible on the local host, you can
do so by submitting the job from the command line. If you plan to use SP-Peptide docking
mode, you must select a grid that was generated for this mode.
If you want to display the receptor in the Workspace, select Display receptor. If you want to see
the grid box, the ligand center box, and the grid center, select Show grid boxes. (The grid boxes
are only displayed for grids generated with the Suite 2011 release onward).
5.3
The Ligands Tab
In the Ligands tab you specify the source of ligands to be docked or scored and set size limits
for skipping ligands. You can also choose the partial charges to use and change the settings for
van der Waals radii scaling of nonpolar ligand atoms.
5.3.1
Specifying the Source of the Ligands
In the Ligands to be docked section you specify the source of ligands to be docked (or scored in
place). The ligands must satisfy the criteria given in Section 3.4 on page 21. To specify the
source, choose one of the following options from the Use ligands from option menu:
• Files—If you select Files as the source of ligands, you can type the file names in the File
name text box, or click the Browse button to open a file selector, in which you can choose
the format and navigate to and select the ligand files. The file must be in Maestro, SD,
MOL2, or PDB format. Maestro and SD files can be in compressed form.
By default, all structures in all files are docked. If you specify a single file, you can select
a contiguous subset of ligands from the file, by entering values in the Range text boxes.
To enter the upper limit, deselect End. The End option allows you to dock all ligands
from a particular ligand to the end of file.
Selecting a subset of ligands is useful if a problem with Glide or with the input ligand
causes a Glide job to fail. Then you can set the initial ligand number when you restart the
Glide 6.7 User Manual
47
Chapter 5: Ligand Docking
job to pick up at the point in the ligand file at which the problem occurred. This only
applies if a single file is selected for docking.
• Project table (selected entries)—Select this option to dock ligands that are selected
entries in the Project Table.
• Workspace (included entries)—Select this option to dock the structures in the Workspace.
If you select this option, the entries in the Workspace must satisfy the criteria for valid
ligand entries.
Figure 5.2. The Ligands tab of the Ligand Docking panel.
5.3.2
Selecting the Source of Ligand Partial Charges
Partial charges for the ligand atoms are obtained from the force field. These partial charges do
not reflect the influence of the environment on the atomic charges. To compensate, you can use
pregenerated partial charges from some other source. To use the partial charges that are
supplied with the ligand, select Use input partial charges.
48
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
The QM-polarized ligand docking solution provides a way of using quantum mechanical
charges from a QSite calculation to dock ligands. For more information, see the document
QM-Polarized Ligand Docking.
5.3.3
Setting Restrictions on the Type of Ligands
In the text boxes at the end of the Ligands to be docked section, you can set limits on structural
feature counts to screen out structures before docking. These text boxes are described below.
Do not dock or score ligands with more than a atoms
Set the maximum number of atoms a ligand structure may have if it is to be docked. Ligand
structures in the input file that have more than the specified number of atoms will be skipped.
The default is also the maximum, 500 atoms. You can reduce the maximum number of atoms
a, if the active-site region is small and enclosed, to speed up a docking calculation.
Do not dock or score ligands with more than r rotatable bonds
Set the maximum number of rotatable bonds a ligand structure may have if it is to be docked
flexibly. Ligand structures in the input file that have more than this number of rotatable bonds
are skipped. The default is also the maximum, 100 rotatable bonds. If only relatively small or
rigid ligand “hits” are wanted, you can decrease the value of r. If you use torsional constraints,
the rotatable bonds that are constrained are excluded from the number of rotatable bonds.
5.3.4
Van der Waals Radii Scaling
Glide does not generally allow for flexible receptor docking, except for reorientation of
selected hydroxyl and thiol groups (see Section 4.5 on page 41).1 However, successful docking
sometimes requires that the ligand and the receptor “give” a bit in order to bind. To model this
behavior, Glide can scale the van der Waals radii of nonpolar atoms (where nonpolar is defined
by a partial charge threshold you can set), thereby decreasing penalties for close contacts.
Scaling is performed for qualifying atoms in the ligand, but not those in the receptor. Ligand
atom radii scaling settings can be changed using the options in this section. To scale receptor
atom radii, you must choose the appropriate options in the Receptor tab of the Receptor Grid
Generation panel prior to grid generation.
Scaling factor
Specify the scaling factor. The default is 0.80. To turn van der Waals radii scaling off, set the
scaling factor to 1.0. Full penalties for close contacts of nonpolar ligand atoms will then be
used.
1.
Flexible receptor docking can be performed with the Induced Fit Docking solution, which uses Glide with
Prime to allow for receptor flexibility. For more information, see the document Induced Fit Docking.
Glide 6.7 User Manual
49
Chapter 5: Ligand Docking
Partial charge cutoff
Scaling is performed only on nonpolar atoms, defined as those for which the absolute value of
the partial atomic charge is less than or equal to the number in the text box. Since this is an
absolute value, the number entered must be positive. The default for ligand atoms is 0.15.
5.4
The Settings Tab
The Settings tab defines the basic options for docking ligands: specifying the grid, selecting
the precision, and setting flexibility options. You can also make settings for the selection of
initial poses and for the energy minimization of the poses that pass the initial selection in the
Settings - Advanced Settings dialog box. To open this dialog box, click Advanced Settings. The
settings, both basic and advanced, are described in the following subsections.
The force field used for docking is the OPLS_2005 force field.
Figure 5.3. The Settings tab of the Ligand Docking panel.
50
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
5.4.1
Selecting the Docking Precision
There are three choices of docking precision, given on the Precision option menu in the
Docking section.
• HTVS (high-throughput virtual screening)—High-throughput virtual screening (HTVS)
docking is intended for the rapid screening of very large numbers of ligands. HTVS has
much more restricted conformational sampling than SP docking, and cannot be used with
score-in-place. Advanced settings are not available for HTVS, as they are predetermined.
• SP (standard precision)—Standard-precision (SP) docking is appropriate for screening
ligands of unknown quality in large numbers. Standard precision is the default.
• XP (extra precision). Extra-precision (XP) docking and scoring is a more powerful and
discriminating procedure, which takes longer to run than SP. XP is designed to be used on
ligand poses that have a high score using SP docking. We recommend that you run your
database through SP docking first, then take the top 10% to 30% of your final poses and
dock them using XP, so that you perform the more expensive docking simulation on
worthwhile ligands. For more information on XP docking, see Section 2.3 on page 13.
• SP-Peptide—Standard-precision docking for peptide ligands uses the same general settings as for regular standard precision but changes some of the settings to enhance the
retention of poses. Specifically, it keeps 100000 poses in the initial docking stage, and
uses 1000 poses per ligand for energy minimization, and sets the number of poses for
post-docking minimization to 100. Although this option was designed for peptides, it is
not restricted to peptides: you can dock any kind of ligand.
If you select this option, the grid must be one that was generated for this mode, i.e. with
Generate grid suitable for peptide docking selected.
If you want to dock a set of ligands using a progression of precision, you can use the Virtual
Screening Workflow to set up and run the docking jobs. See the Virtual Screening Workflow
document for details.
5.4.2
Setting Ligand Sampling Options
From the Ligand Sampling option menu you can choose whether ligands are docked flexibly or
rigidly, refined in place, or simply scored in place; and set options for conformation generation
for flexible docking. The choices are described below.
Flexible
This is the default option, and directs Glide to generate conformations internally during the
docking process; this procedure is known as flexible docking. Conformation generation is
limited to variation around acyclic torsion bonds and pyramidal nitrogen inversions such as in
Glide 6.7 User Manual
51
Chapter 5: Ligand Docking
amines and sulfonamides. For a set of predefined functional groups, such as amides and esters,
you can bias sampling of the torsion around the bond that normally adopts a particular conformation so that it adopts the desired conformation. These variations are controlled by the
following options:
• Sample nitrogen inversions—Sample inversions at pyramidal nitrogen atoms. This option
is on by default.
• Sample ring conformations—Sample ring conformations, using the same technology as in
LigPrep—see Section 4.11 of the LigPrep User Manual for details. This option is
selected by default. Only low-energy conformations are kept. You can choose to include
the original ring conformation by selecting Include input ring conformation. (This conformation is of course used if you do not sample ring conformations.)
• Bias sampling of torsions for—Choose an option for sampling of torsions that should normally be restricted to a particular conformation. The biasing can include retention of the
input conformation, setting the torsion to a particular value, applying a penalty for deviating from the desired conformations, or allowing only a particular conformation to within
a small angle range. The options cover different selections of functional groups:
• Predefined functional groups—Bias the sampling of torsions for a set of functional
groups that is defined in a resource file. The choice of biasing method, as outlined
above, is set in the resource file. The resource file can be customized—see
Section 7.6.2 on page 111.
• Amide bonds—Bias the sampling of torsions around the C–N bond of amides only.
You can choose the biasing method from the option menu:
• Penalize nonplanar conformation—Apply penalties to amide bonds that are not
cis or trans, rather than freezing them entirely.
• Retain original conformation—Fix amide bonds in their input conformation.
• Allow trans conformation only—Enforce the trans conformation to within a
small angle range (20° by default). Ligands that do not dock in a trans conformation are rejected.
The default is to penalize nonplanar conformations.
• None—Do not apply any biasing, but allow the conformation to be determined by
the conformation generator via the force field.
Rigid
Rigid docking does not generate conformations of the input ligand. It only allows the input
conformation of the ligand to be translated and rigidly rotated relative to the binding site.
52
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
None (refine only)
When this option is selected, Glide does no docking, but rather uses the input ligand coordinates to perform an optimization of the ligand structure in the field of the receptor, and then the
ligand is scored. The goal of this docking method is to find the best-scoring pose that is
geometrically similar to the input pose. For HTVS and SP, a minimization is performed; for
XP, the ligand is regrown in place. This option is not a substitute for a full docking calculation,
and requires accurate initial placement of the ligand with respect to the receptor.
None (score in place only)
When this option is selected, Glide does no docking, but rather uses the input ligand coordinates to position the ligands for scoring. It therefore requires accurate initial placement of the
ligand with respect to the receptor.
This option is useful to score the native ligand in its cocrystallized or modeled position, or as a
post-processing step on Glide-generated poses to obtain individual components of the GlideScore prediction of the binding affinity. It can also be used to check whether the scores of the
known binders in their native proteins are similar enough to their scores when cross-docked to
the chosen receptor protein. If this is the case, it is reasonable to expect that similar structures
would also score well.
Score in place should not be used with Glide XP, as full XP sampling is normally needed to
avoid strong XP penalties for ligands that should be able to dock correctly.
Note:
5.4.3
You cannot use score in place for the ligand that is defined as the reference ligand for
calculation of the RMSD in conformational comparisons.
Writing XP Descriptor Information
If you want to visualize the various scoring terms from the XP scoring function in the XP Visualizer panel, select Write XP descriptor information. These terms are written to a file that can be
read by the XP Visualizer. You can write these terms only for XP docking runs.
Note:
5.4.4
To generate XP descriptor information, you must have a license for this feature.
Adding Epik State Penalties to the Docking Score
If the ligands were prepared using Epik for ionization and tautomerization, the Epik penalties
for adopting higher-energy states (including those where metals are present) are added to the
docking score when the Add Epik state penalties to docking score option is selected. Ligands
that do not have this information are not penalized and will therefore have better scores, so you
should ensure that the ligand set is consistent. This option is selected by default.
Glide 6.7 User Manual
53
Chapter 5: Ligand Docking
If the ligand interacts with a metal (distance less than 3.0 Å), the metal penalties that are
computed when Epik is run with the metal binding option are used. If multiple ligand atommetal interactions are found, the smallest value of the metal-specific penalty is used.
5.4.5
Rewarding Intramolecular Hydrogen Bonds
Ligands with intramolecular hydrogen bonds pay a smaller entropic penalty upon binding, so
forming intramolecular hydrogen bonds can be important for binding. You can add a reward
for each intramolecular hydrogen bond to the GlideScore, by selecting Reward intramolecular
hydrogen bonds. A reward is also added to Emodel for each intramolecular H-bond, to favor
selection of poses with intramolecular hydrogen bonds.
5.4.6
Improving Planarity Around sp2 Atoms
Groups such as aromatic rings, amides, and esters can adopt a geometry that is nonplanar. This
happens because the torsional potential has a finite barrier which can be overcome to some
extent by other interactions, and is a physically reasonable effect. In Glide docking, nonplanarity can also be a result of the approximations made to reduce clashes with the receptor.
Selecting Enhance planarity of conjugated pi groups increases the torsional potential around
bonds between atoms whose geometry should be planar (i.e. sp2 atoms). This option should
make aromatic rings, amides, esters, and so on, less likely to adopt a nonplanar geometry.
To some extent, planarity can be enforced in flexible docking by choosing one of the biasing
options for sampling torsions. However, if you want to improve planarity in post-docking minimization or for non-flexible docking, you should select this option.
5.4.7
Applying Excluded Volumes
If the grid that you chose contains excluded volume data, you can choose to apply the excluded
volumes when docking the ligands, and set the penalty for occupying the excluded volume. To
do so, select the option Apply size excluded volume penalties, and choose a size for the penalty.
The sizes available are large, medium, and small, which correspond to particular choices of the
penalty value. The penalty ramps up linearly from zero at the edge of the sphere to the value
specified, over a shell of width 10% of the sphere radius, and remains at that value over the rest
of the sphere. Each atom that occupies the excluded volume incurs the penalty, and the penalty
is applied in both the rough scoring stage and the final docking. You can also view the
excluded volumes in the Workspace, by selecting Show excluded volumes. The excluded
volumes controls are not available if the grid does not have excluded volumes.
54
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
5.4.8
Advanced Settings
Further options for controlling the docking process are available in the Settings - Advanced
Settings dialog box, which opens when you click Advanced Settings in the Settings tab. You
will not normally need to change these options from their defaults.
Figure 5.4. The Settings - Advanced Settings dialog box.
5.4.8.1
Conformer Generation
When you sample ring conformations, only the low-energy ring conformations are kept. The
threshold for keeping or discarding ring conformations can be set in the Conformer generation
section of the dialog box. Ring conformations are discarded if their energies are higher than
that of the lowest conformation by more than the amount specified in the Energy window for
ring sampling text box.
The poses obtained in docking can depend on details of the input structure. Sometimes this is
useful, to obtain a wider range of poses, but it can also prevent comparisons between docking
runs that should not depend on the detailed structure of the input ligands. To remove the dependency on the input structure, you can select Canonicalize input conformation. The input coordinates of the ligand are then discarded and regenerated, based only on the connectivity
(including bond orders) and the stereochemistry.
Glide 6.7 User Manual
55
Chapter 5: Ligand Docking
While variation in docked poses with the input structure can be a problem, it can also be used
to enhance conformational sampling. You can add variations to the input structure that are used
as input to the conformational sampling, by selecting Use enhanced sampling, and specifying
the enhancement factor in the Enhance conformational sampling by N times text box. This
factor is the number of structures (including the input structure) that are used as seed structures
for conformational sampling.
5.4.8.2
Selection of Initial Poses
The options in the Selection of initial poses section of the dialog box control the way poses pass
through the filters for the initial geometric and complementarity “fit” between the ligand and
receptor molecules. The grids for this stage contain values of a scoring function representing
how favorable or unfavorable it would be to place ligand atoms of given general types in given
elementary cubes of the grid. These cubes have a constant spacing of 1 Å. The “rough score”
for a given pose (position and orientation) of the ligand relative to the receptor is simply the
sum of the appropriate grid scores for each of its atoms. By analogy with energy, favorable
scores are negative, and the lower (more negative) the better.
The initial “rough scoring” is done on a coarse grid, on which the possible positions for placing
the ligand center are separated by 2 Å, twice the elementary cube spacing, in x, y, and z. The
“refinement” step rescores the successful rough-score poses after the particular rigid translational repositioning of –1.0 Å or +1.0 Å in x, y, and z that gives the best possible score. This
procedure effectively doubles the resolution of the scoring screen.
Poses that pass these initial screens enter the final stage of the algorithm, which involves evaluation and minimization of a grid approximation to the OPLS-AA nonbonded ligand-receptor
interaction energy.
This section contains three text boxes and an option for controlling the initial screening:
Keep n initial poses per ligand for the initial phase of docking
This text box sets the maximum number of poses per ligand to pass to the grid refinement
calculation. The value must be a positive integer. The default setting depends on the type of
docking specified and whether Glide constraints have been applied:
•
•
•
•
56
5000 poses for flexible docking jobs in general
500 poses for flexible docking jobs to which Glide constraints are applied
1000 poses for rigid docking jobs (Glide constraints do not change this value)
100000 poses for peptide docking jobs
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
Scoring window for keeping initial poses
This text box sets the rough-score cutoff for keeping initial poses, relative to the best rough
score found so far. The value must be positive. The default window is 100.0 kcal/mol, meaning
that to survive, a pose must score within 100.0 kcal/mol of the best pose. Using the default
window, for example, if the best pose found so far has a score of -60.0 kcal/mol, poses with a
score more positive than 40.0 kcal/mol are rejected.
Keep best m poses per ligand for energy minimization
This text box specifies the number of poses per ligand to be energy minimized on the OPLSAA nonbonded-interaction grid. The default setting depends on the type of docking specified:
•
•
•
•
400 poses for flexible SP docking jobs, 800 poses for flexible XP docking jobs.
40 poses for flexible docking jobs to which Glide constraints are applied.
100 poses for rigid docking jobs (Glide Constraints do not change this value).
1000 poses for peptide docking jobs.
The range for this value is 1 to n, where n is the value in the Keep n initial poses per ligand for
the initial phase of docking text box, with the exception that for XP docking, the number can be
adjusted between 800 and 1000.
Use expanded sampling
This option bypasses the elimination of poses based on the rough score, so that many more
poses are passed on to the subsequent stages. This option is particularly useful for fragment
docking to ensure that good poses are found.
5.4.8.3
Energy Minimization Settings
The energy minimization stage of the docking algorithm minimizes the energy of poses that
are passed through the Selection of initial poses scoring phase. The Energy minimization section
of the Settings - Advanced Settings dialog box contains two options:
Distance-dependent dielectric constant
Glide uses a distance-dependent dielectric model in which the effective dielectric “constant” is
the supplied constant multiplied by the distance between the interacting pair of atoms. The
default setting is 2.0, and Glide’s sampling algorithms are optimized for this value. Although
this text box allows you to set the dielectric constant to any real value greater than or equal to
1.0, changing this setting is not recommended. This relatively strong dielectric is sometimes
needed to “hold” a hydrogen bond in the energy-grid optimization phase of the algorithm.
Glide 6.7 User Manual
57
Chapter 5: Ligand Docking
Maximum number of minimization steps
This text box specifies the maximum number of steps taken by the conjugate gradient minimization algorithm. The value must be greater than or equal to 0; the default value is 100. Setting
the value to 0 results in a single-point energy calculation on each pose that survives roughscore screening, or on the single initial pose if no screening was done.
5.4.8.4
Hydrogen Bonds with Aromatic Hydrogens and Halogens
In addition to normal hydrogen bonds, you might want to account for other noncovalent interactions that are generally weaker, but could be important. You can choose to include three
types of interactions in your docking run: hydrogen bonds to aromatic hydrogens, hydrogen
bonds to halogens (halogen as an acceptor), and halogen bonds (halogen as a donor). The
options to include these interactions are in the Count aromatic H and halogen H-bonds section.
(Strictly speaking, halogen bonds aren’t H-bonds, but they are included here as a class of weak
noncovalent interactions.) Currently you can only choose to have halogens treated as donors or
as acceptors, but not both. These interactions contribute to the scoring and constraints in the
same way as normal hydrogen bonds.
For aromatic hydrogens, you can set a cutoff for the partial charge, to filter out hydrogens that
have small partial charges. To be considered as a donor, the partial charge on an aromatic
hydrogen must be greater than the value given in the Partial charge cutoff text box.
5.5
The Core Tab
In the Core tab, you can specify the “core” of a ligand, and use this core to constrain the
docking of other ligands or to calculate the rms deviation from the defined core for all docked
ligands. The core is specified in terms of a reference ligand.
The constraint that you apply here is a ligand-based constraint, which means that the ligands
that are subject to the constraint are those that have the same core moiety as the reference
ligand. Ligands that do not match the core can be screened out in the first docking stage. In
later stages, ligands that match the core pattern but do not meet the rms tolerance for the position of the core can be screened out. Receptor-based constraints can be applied in the
Constraints tab, which is described in the next section.
58
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
Figure 5.5. The Core tab of the Ligand Docking panel.
To use a core definition, select Use core pattern comparison, then select an option for how to
use the core definition:
• Use for RMSD calculations only—Use the core definition for computing the RMSD from
the reference ligand, but not for restricting docking.
• Restrict docking to reference position—Restrict the docking of ligands so that the ligand
core lies within a given RMSD of the core in the reference ligand. The core is defined in
terms of a set of atoms or a SMARTS pattern; if the ligand does not contain the topology
defined in the core, it can be skipped. You can enter a tolerance for the RMSD in angstroms for restricting the docking to the reference position in the Tolerance text box.
The Core tab offers a great deal of flexibility in defining the core. You can use the entire molecule (or its heavy atoms) for the RMSD evaluation, or you can define a subset of the molecule,
both for the matching and for the RMSD, in terms of SMARTS patterns. Choosing the core
molecule and setting up the core pattern is done in the Define core section of the Core tab. The
Glide 6.7 User Manual
59
Chapter 5: Ligand Docking
controls in this section become available when you choose Use for RMSD calculations only or
Restrict docking to reference position.
If you plan to use the entire reference ligand for the core (or its non-hydrogen atoms), ensure
that Pick core-containing molecule is selected, then pick an atom in the Workspace that belongs
to the core molecule. The molecule is marked in purple in the Workspace if Show markers is
selected. Pick core-containing molecule is automatically selected when the controls in this
section become available. You can then proceed to choose All heavy atoms or All atoms for the
core, under Core atoms.
If you want to define the core in terms of a smaller set of atoms, select SMARTS pattern under
Core atoms. Pick core-containing molecule is automatically deselected if you make this choice,
because you can choose any atoms and picking a molecule is no longer relevant. You can then
enter a SMARTS pattern into the text box, or you can select atoms in the Workspace with the
Select atoms tool,
then click Get From Selection to create a SMARTS pattern for the selected atoms. The atoms in
the core molecule that match the pattern are marked in the Workspace with green markers if
Show markers is selected. Note that it is advisable not to have the receptor displayed when you
click Get From Selection, because it can slow the generation of the SMARTS pattern.
You might not want to use all the atoms in the SMARTS pattern for the RMSD calculation. If
this is the case, you can choose Pick RMSD subset atoms, and pick atoms in the Workspace
that are in the SMARTS pattern to define a subset of atoms for the RMSD calculation. A lock
symbol appears next to the atoms you pick. Each pick adds to the set; picking an atom again
deletes it. The RMSD atoms are used both when calculating only the RMSD and when
restraining the core: it is the RMSD of these atoms that must fall below the prescribed tolerance in the latter case.
Finally, you can select Skip ligands that do not match core pattern if you want to strictly apply
the core pattern filter. This option is selected by default. When it is selected, ligands that do not
match the core are discarded in the first stage of the funnel. If this option is not selected,
ligands that do not match the core pattern are docked anyway, the core constraints are not
applied, and the RMSD is only calculated for those ligands that do match the core pattern.
60
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
5.6
The Constraints Tab
The Constraints tab lists all the Glide constraints that are defined for the receptor grid file you
specified in the Settings tab, and provides the means to apply these constraints in docking. The
available constraints are shown in the Workspace if Show markers is selected and the receptor
is displayed. To display the receptor, select Display Receptor. Positional and NOE constraints
are marked with gray spheres. Atoms for which H-bond/metal constraints are defined are
marked with a red asterisk.
Constraints are applied during or after docking by identifying the relevant features in the
ligand, and requiring specified atoms in those features to be spatially confined to the constraint
region. The ligand features and the atoms in those features that must be constrained are defined
in terms of SMARTS patterns. You can customize the features in the Edit Feature dialog box,
which is described in Section 5.6.2 on page 64. For positional constraints, no default features
are defined, so you must provide a feature definition in the Edit Feature dialog box.
Figure 5.6. The Constraints tab of the Ligand Docking panel.
Glide 6.7 User Manual
61
Chapter 5: Ligand Docking
Glide constraints can be used in HTVS mode, but may result in poor pose recovery relative to
unconstrained HTVS docking.
Only the constraints you select will be applied in the ligand docking job you are setting up. If
there are no constraints selected when you start the docking job, no constraints will be applied.
5.6.1
Setting Constraints
To allow some flexibility in setting constraints, including optional constraints, Glide provides
four constraint groups, presented in subtabs labeled Group 1 through Group 4. If you want to
set simple constraints you can do so in the Group 1 tab, which is displayed by default, and
ignore the remaining groups.
Each subtab contains a table of all the available constraints (described in Table 5.1), an Edit
Table 5.1. Description of the Available constraints table.
Column
Description
Use
Check box to select the constraint for use in docking. Click to select or deselect.
Name
Constraint name. This is the name defined in the Constraints tab of the Receptor
Grid Generation panel.
Receptor
Constraint Type
Type of receptor constraint. Hydrogen bond constraints are classified into H-bond
for hydrogen-bond acceptors and Polar Hydrogen for hydrogen-bond donors (i.e.
hydrogen atoms).
Ligand Feature
Name of the feature in the ligand that must match the constraint. The available features are: Acceptor, Charged Acceptor, Neutral Acceptor, Donor, Hydrophobic,
and Custom. By default, the feature that matches a receptor polar hydrogen is set
to Acceptor; for a receptor H-bond type it is Donor, for a receptor hydrophobic
feature it is Hydrophobic, and for a positional constraint it is Custom. Custom is
undefined by default, so you must edit this feature to define the patterns that match
the desired ligand atoms.
Number of
Ligand Atoms
In this column, the number of ligand atoms that must occupy a hydrophobic region
can be set. The value is only meaningful for hydrophobic constraints.
Feature button, and two Must match options. To set constraints within a tab, first click the Use
column for each of the desired constraints. An X appears in the check box. If the receptor is
displayed and Show markers is selected, the hydrophobic and positional constraint markers
turn red when the constraint is selected, and the H-bond and metal constraint markers turn
turquoise. The number of constraints you selected is displayed next to the table title.
Next, decide whether all the constraints should be applied, or only some of them.
62
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
• If you want all of the chosen constraints to be applied, select All under Must match.
• If you want some of the chosen constraints to be applied, select At least under Must
match, and enter the number of constraints that must be satisfied in the text box. For
example, if you chose three hydrogen-bond acceptors, and you want any two out of the
three to be satisfied, you would enter 2 in the text box.
The number of constraints that are required is displayed in the subtab tab, and is added to the
total, which is displayed in the upper part of the Constraints tab.
To clear a constraint, click the Use column again. The X in the check box disappears.
More complex constraints can be applied by using more than one constraint group. If you set
constraints in more than one group, each group of constraints is applied to the ligand (that is, a
Boolean AND is applied between groups). The general syntax of the constraints is (N1
required from Group 1) AND (N2 required from Group 2) AND (N3 required from Group 3)
AND (N4 required from Group 4).
For example, suppose the grid contains a hydrogen bond constraint hbond1 and two positional
constraints pos1 and pos2. If you want to enforce the hydrogen bond constraint, and require
only one of the two positional constraints, you would use two groups. In Group 1, you would
click the Use check box for hbond1 and select All under Must match. In Group 2, you would
click the Use check boxes for pos1 and pos2, select At least under Must match, and enter 1 in
the text box. Both groups are applied. The resulting constraints can be represented as “hbond1
AND (pos1 OR pos2)”.
If your desired constraint specification cannot be put in the general form above, you might be
able to achieve your goal by running more than one docking job with a separate constraint
specification for each. For example, if you want to apply the constraints “(hbond1 AND pos1)
OR (hbond2 AND pos2) OR (hbond3 AND pos3)”, you could run three separate docking jobs,
one with (hbond1 AND pos1) set, one with (hbond2 AND pos2) set, and one with (hbond3
AND pos3) set.
The total number of required constraints, summed over all groups, must be four or fewer. The
number in parentheses after the group name in the subtab tab is the number of constraints that
must be satisfied in this group. If you selected All, it is the number of constraints you chose. If
you selected At least, it is the value in the At least text box.
A default ligand feature definition is supplied and assigned to each receptor constraint type,
with the exception of positional constraints, for which the Custom feature is undefined. To
change either the assignment or the feature definitions, use the Edit Feature dialog box, as
described in the next section. You can select the same constraint in more than one group, but
the assignment and the feature definitions are the same for each group.
Glide 6.7 User Manual
63
Chapter 5: Ligand Docking
To use positional constraints, you must define the ligand feature that it should match. No
default feature definition is provided, and the job cannot be run until a feature definition is
provided. To define the ligand feature, select the appropriate row in the Available constraints
table and click Edit Feature. The feature can then be defined in the Edit Feature dialog box.
5.6.2
Defining Ligand Features
Ligand features are identified by matching to a collection of SMARTS patterns that defines a
feature type. The available feature types are:
• Acceptor—Neutral or charged acceptor.
• Charged Acceptor—Acceptor that has a formal charge. Includes O• , S• , F• , Cl• . Both
oxygens in a carboxylate are included as charged acceptor atoms.
• Neutral Acceptor—Acceptor that does not have a formal charge. Includes a range of
groups containing N and O.
• Acceptor Including Halogens—Acceptor or a neutral halogen (F, Cl, Br, I).
• Donor—Hydrogen attached to any neutral nitrogen, oxygen or sulfur; also includes OH•
• Donor Including Aromatic H—Donor as defined above and any hydrogen attached to an
aromatic carbon.
• Donor Including Halogens—Donor as defined above and any neutral halogen other than F.
Halogens can make nonbonded interactions with acceptor groups.
• Donor Including Aromatic H + Halogens—Donor, hydrogen attached to aromatic carbon,
or halogen, as defined above.
• Hydrophobic—A range of patterns of aliphatic and aromatic carbons, and neutral halogens other than F.
• Custom—Specify your own feature type by defining SMARTS patterns that match the
feature.
The donor or acceptor feature type is coordinated with the settings for aromatic and halogen Hbonds in the Settings - Advanced Settings dialog box (see Section 5.4.8.4 on page 58). A
change in either place changes the setting in the other.
The feature definitions for these types form a feature set, which can be imported and exported.
Each constraint can have its own feature definition, so you can have a different definition of a
given feature type for each constraint. However, the same feature definition from the same set
is used for a given constraint in all groups. For each feature definition, you can add patterns,
edit and delete custom patterns, and define patterns for exclusion of functional groups. Feature
64
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
sets can be imported and exported. These tasks are carried out in the Edit Feature dialog box,
which you open by clicking Edit Feature.
Figure 5.8. The Edit Feature dialog box.
To change the feature type assigned to the constraint or to edit the feature definition for a
particular constraint, select the row in the Available constraints table, and click Edit Feature.
If you only want to change the feature type assigned to the constraint, choose the feature type
from the Feature menu, and click OK. The change is reflected in the Ligand Feature column of
the Available constraints table, and applies to all groups. If you want to edit the feature definition, follow the instructions in Section 5.6.2.2 and Section 5.6.2.3.
5.6.2.1
Loading and Saving Feature Sets
Built-in feature sets are stored with the distribution, so you do not need to create your own,
except for positional constraints.
You can import a feature set for the selected constraint from a file by clicking Import, and navigating to the feature file. When you import a feature set, the definitions of all feature types are
replaced, not just the feature type chosen from the Feature menu. The feature definitions are
replaced only for the selected constraint, but are replaced for that constraint in all groups.
Feature sets can be saved to a file by clicking Export, and specifying the file location in the file
selector that is displayed.
5.6.2.2
Adding, Editing, and Deleting Patterns
The patterns that define a feature are displayed in the Pattern list table (see Table 5.2) when you
choose the feature type from the Feature option menu.
Glide 6.7 User Manual
65
Chapter 5: Ligand Docking
If the patterns in a given feature definition do not cover all the functional groups that you want
to include in the definition, you can add extra patterns. Matching of patterns to ligand structures is done in the order specified in the Pattern list table. You cannot change the order of the
patterns once they are in the table, so you must add new patterns at the appropriate point. If you
want to move a pattern, you must delete it and add it again.
To add a new SMARTS pattern, click the table row above which you want the pattern to be
inserted, then click New. In the New Pattern dialog box, you can provide a SMARTS pattern
and define the atoms that must satisfy the constraint: the acceptor or donor atoms, for example.
Figure 5.10. The New Pattern dialog box.
There are two ways to provide a SMARTS string. The first is to type the string into the
SMARTS pattern text box. The second is to select atoms in the Workspace, then click Get from
selection. Maestro generates a SMARTS string for the selected atoms and places it in the
SMARTS pattern text box. You can then edit it if you like. You should make sure that you have
a suitable molecule, such as a known active ligand, in the Workspace before you open the Edit
Feature dialog box. This second procedure can be slow if large molecules, such as a receptor,
are displayed in the Workspace.
Table 5.2. Description of Pattern list table.
Column
Description
Mark
Column of check boxes. Selecting a check box marks the pattern on any structures
that are displayed in the Workspace.
Pattern
Pattern definition. The definitions are SMARTS strings.
Atoms
The list of atoms that may be constrained, numbered according to the SMARTS
string.
Exclude
Column of check boxes. If a check box is selected, the atoms in other patterns are
matched only if they do not match this pattern. This is essentially a NOT operator.
Excluded patterns are processed before other patterns.
To specify the atoms in the ligand that must satisfy the constraint, enter the atom numbers as a
comma-separated list in the Atom numbers text box. The atom numbers are given relative to
the SMARTS pattern: atom 1 is the first atom in the SMARTS pattern, and so on.
66
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
For positional, metal, and H-bond constraints, you can specify multiple atoms, but only one of
the specified atoms is constrained. Normally, only one atom should be specified, to ensure that
the correct atom is constrained. However, if the functional group contains multiple acceptors or
donors, for example, you can specify all of them if you want any one of them to be constrained.
This is not necessary for groups such as carboxylates, where the local symmetry of the functional group is used and either of the oxygen atoms in the carboxylate can be selected for the
constraint, even though only one of them is specified.
For hydrophobic constraints, the non-hydrogen atoms of the hydrophobic group should be
specified. The number of atoms that is constrained is specified in the Required Ligand Atoms
column of the Available constraints table.
When you click OK, a new row is added to the Pattern list table.
To edit a pattern, select the table row for the pattern, then click Edit. In the Edit Pattern dialog
box, you can modify the SMARTS pattern or get a new pattern from the Workspace selection,
and change the atoms in the pattern that must satisfy the constraint. The Edit Pattern dialog box
is identical to the New Pattern dialog box except for the title.
To delete a pattern, select the table row for the pattern, then click Delete.
5.6.2.3
Excluding Functional Groups from a Feature
If you want to ensure that certain functional groups are not matched, you can select the check
box in the Exclude column for the pattern for that group. For example, you might want to
exclude a carboxylic acid group from being considered as a hydrogen bond donor, because it
will be ionized under physiological conditions. Excluded patterns are processed first, regardless of their position in the Pattern list table. Thus, an excluded pattern prevents the atoms that
match it from being matched by any other pattern.
5.6.2.4
Visualizing Patterns
If you want to see a pattern for a given ligand or group of ligands, you can select the check box
in the Mark column for the pattern. Any occurrences of the pattern are marked in the Workspace. You can display markers for more than one pattern, but the markers do not distinguish
between patterns. If you want to see the atoms and bonds as well as the markers, select Apply
marker offset.
5.6.3
Choosing When To Apply Constraints
Constraints can be applied during the docking process, or only after the ligands have been
docked. When testing for constraint satisfaction after docking, constraints are used as a postdocking filter and have no influence on the docking or scoring. You can therefore compare
results from such a run directly with an unconstrained docking run. To apply constraints only
Glide 6.7 User Manual
67
Chapter 5: Ligand Docking
after docking, select Test constraint satisfaction only after docking. This option applies only to
the constraints available in the Constraints tab, not to core constraints or torsional constraints.
5.6.4
Using Multiple Constraints
When you use multiple constraints with complex optional constraint specifications and many
ways of satisfying the constraints, it is possible to lose poses because Glide cannot track all
possible ways a given pose can satisfy the constraints, due to memory limitations. Under these
circumstances, Glide truncates the list of poses or the selections of constraint-satisfying atoms
per pose, or both. The result is that poses that would ultimately satisfy the constraints are lost.
There are several ways in which you can alleviate this problem:
• Select Test constraint satisfaction only after docking in the Constraints folder when setting
up the docking job. (If you run from the command line, add the keyword finalonly
after constraints in the jobname.inp file.) This avoids the memory explosion needed
to track complicated constraint combinations through the Glide funnel, but still ensures
that the final docked poses satisfy the required constraints.
• Use more restrictive SMARTS patterns to define the kinds of ligand atoms that can satisfy the constraints. This can reduce the memory requirements for tracking constraint satisfaction in a given ligand to a point where all constraint-satisfying poses can be kept. If
you need to allow more general chemistry to satisfy your constraints, you can run multiple Glide jobs to cover the various possibilities.
• Split your job into several jobs with tighter restrictions on which constraints must be satisfied, while still spanning all the desired possibilities. For instance, suppose you want
your ligands to make hydrogen bonds to each of three protein hydroxyl groups, but don’t
care which molecule is the donor and which is the acceptor for each group. A single job
to cover all the possibilities would have three groups of two constraints each, each group
consisting of H-bond constraints to the hydrogen and oxygen of one of the protein
hydroxyls, and would require one of the two constraints in each group to be satisfied.
This job is likely to run into the memory problems and lose poses, particularly for ligands
that have multiple H-bonding groups (e.g. sugars). Instead, you could run eight Glide
jobs, one for each specific combination of donors and acceptors among the three protein
hydroxyls. Each job would specify a single group of three constraints, all required, and
would be less likely to lose poses due to memory restrictions.
68
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
5.7
The Torsional Constraints Tab
There are situations in which you want to constrain some of the torsional degrees of freedom in
the ligand. For example, a ligand in the binding site might have only one conformation of a
particular rotatable group, while other groups can exist in several conformations. Or the ligand
might have a large number of rotatable bonds, such as in a polypeptide. Glide provides a means
of constraining selected torsional degrees of freedom in the Torsional Constraints tab. The
groups that are constrained are defined in terms of SMARTS patterns, and you can constrain all
of the torsions in the group or only selected torsions in the group.
This tab is only available when either Dock flexibly or Refine is selected in the Settings tab.
Figure 5.11. The Torsional Constraints tab of the Ligand Docking panel.
Glide 6.7 User Manual
69
Chapter 5: Ligand Docking
To set up and apply the constraints, you need to define the rotatable groups, and for each group,
the torsions to constrain. The easiest way of defining the groups is to include a ligand in the
Workspace and select the atoms in the group with the Select atoms tool:
When you have selected the atoms, click New. The New Torsion Pattern dialog box opens, and
you can click From Workspace Selection to get the SMARTS pattern from the atoms that you
selected in the Workspace. You can also type in a SMARTS pattern directly. When you click
OK, a new row is added to the Patterns table that shows the SMARTS pattern the number of
atoms in the pattern and the number of torsions defined (which is initially zero). The pattern is
marked on the matching atoms in the Workspace.
The SMARTS pattern cannot be changed. If you want to modify it, you will have to create a
new pattern, and delete the old one.
To delete a pattern, select it in the Patterns table, and click Delete. To delete all patterns, click
Delete All.
Once you have entered a SMARTS pattern, you can then select it in the table to define the
torsions that are constrained within that pattern. If you want to constrain all the torsions for this
pattern, select All torsions. The Torsions column in the Pattern table then shows All. If you want
to choose the torsions to constrain, select Specified torsions, and then you can do so by picking
atoms.
1. Select Pick atoms to add a torsion.
2. Pick four atoms in the Workspace to define the torsion.
The atoms must form a contiguous set of three bonds when you have finished picking,
and must not define a ring torsion. They must also be atoms in the SMARTS pattern for
which you are defining torsions.
When you have picked the torsion, the atom numbers are shown in the table. The torsion is
marked in the Workspace with the number in the table and the current value of the angle, and
the number of torsions in the Torsions column of the Patterns table is updated.
No checking is done for the validity of the torsion prior to docking, so you must make sure that
it meets the criteria given above. The torsion cannot be changed once you have defined it, so if
it is in error, you must delete it.
To delete a torsion, select it in the table and click Delete.
See Section 3.11 of the Glide Quick Start Guide for an example of setting and using torsional
constraints.
70
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
5.8
The Output Tab
The options in the Output tab control the final output of ligand poses that pass successfully
through Glide’s various scoring stages. The features of this tab are options in a section called
Structure output and an Advanced Settings button. Clicking the button opens the Output Advanced Settings dialog box.
Figure 5.12. The Output tab of the Glide panel.
5.8.1
Structure Output Options
In flexible docking runs, Glide appends docked poses to a file named jobname_raw.suffix,
where suffix depends on the output file format (.maegz or.sdfgz). The “raw” in the filename
indicates that these poses are not sorted by GlideScore. Once the Glide job has docked all the
ligands, it runs the glide_sort utility script to sort the raw poses and write them to
jobname_pv.suffix or jobname_lib.suffix. The jobname_raw.suffix file is then deleted.
Glide 6.7 User Manual
71
Chapter 5: Ligand Docking
Rigid docking runs sort the poses internally, then, at the end of the job, the sorted poses are
written to either jobname_pv.suffix or jobname_lib.suffix. The glide_sort utility is not
automatically used in rigid docking jobs.
For score-in-place jobs, the structure output file contains the input structures, since they are not
altered.
The output files for the job are written to a subdirectory of the current directory, which is
named with the job name.
The glide_sort utility can sort any pose file, whether _raw.suffix, _pv.suffix, or
_lib.suffix, using default or alternate criteria—see Section 7.5.1 on page 109.
The options in the Structure output section are described below.
File type
The final list of poses that pass Glide’s criteria are written to a multi-structure compressed pose
file. You can select one of two options for the file type:
• Pose viewer file (includes receptor)
This option includes the receptor and the poses in the output pose file, and is intended for
use with the Pose Viewer. It is in Maestro format and is named jobname_pv.maegz.
• Ligand pose file (excludes receptor), in format format
This option includes only the ligand structures in the output pose file. This file cannot be
used with the Pose Viewer, but may be useful if the output poses are intended for input to
a subsequent Glide job, or for some other purpose. It is named jobname_lib.suffix,
where suffix is determined by the choice of output format made from the option menu.
The choices are Maestro and SD, resulting in a .maegz or an .sdfgz suffix.
You cannot write ligand pose files with XP descriptors, with per-residue interactions, or if you
have rotatable groups in the grid. These features require a pose viewer file.
Limit the number of poses to report
This option limits the total number of the predicted best-binding poses written to the sorted
pose file to the specified value. The default is to report all poses that pass all the filters.
Write out at most m poses per ligand
This text box limits the number of poses per ligand written to the sorted pose file. The default
choice of 1 pose per ligand is intended for database screening applications. A larger choice
may be appropriate for lead-optimization studies or whenever several “reasonable” poses are
wanted; for example, to generate a variety of docked poses for study with Liaison or another
72
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
post-docking program. However, if you have a small binding pocket or ligands with few rotatable bonds, choosing a larger number might simply retain poor poses.
Compute RMSD to input ligand geometries
This option enables the computation of the RMSD between the docked poses and the corresponding input ligand geometries. The RMSD calculation is done in place with heavy atoms
only. The value is recorded as a Maestro property named glide rmsd to input (internal name
r_i_glide_rmsd_to_input).
Write report file
A report file in plain text format that lists information about the scoring of the retained poses is
written to jobname.rept if this option is selected. Like the pose viewer file or ligand pose
file, the report file lists the poses in rank order by the selected final-scoring function. The
report file is not written for distributed jobs.
For score-in-place jobs, instead of the .rept file, a .scor file is produced, containing scoring
information for the input structures.
5.8.2
Post-Docking Minimization
The docking process relies on rapid generation of ligand conformers and use of a grid to represent the receptor. The ligand poses generated during docking are rarely exactly at a local
minimum, and post-minimization can improve the geometry of the poses. The Grid Minimization step of the Glide funnel (see Figure 2.2 on page 9) does not perform a full force-field
minimization and excludes interactions beyond 1,5 interactions. The full force-field minimization performed by post-docking minimization penalizes highly strained ligand geometries and
eliminates poses with eclipsing interactions, many intraligand close contacts, and so on.
The Perform post-docking minimization option specifies a full force-field minimization of topranked poses following the final docking. This minimization optimizes bond lengths and
angles as well as torsional angles, and rescores the poses using the scaled Coulomb-van der
Waals term and the GlideScore. You can perform the post-docking minimization for both flexible and rigid docking.
When you perform post-docking minimization, you can select the number of poses per ligand
to minimize. Testing has indicated that the rescoring of poses after post-docking minimization
generally finds a lower GlideScore than is reported in the top few poses from docking. Thus, it
is strongly recommended to apply the minimization to a number of poses. You can enter the
number in the Number of poses per ligand to include text box. This number must not be less
than the number of poses per ligand that is written to the output file. The default is 10 for XP, 5
for SP or HTVS, and 100 for SP-Peptide. If the binding pocket is small, or the ligands are
Glide 6.7 User Manual
73
Chapter 5: Ligand Docking
fairly rigid, the post-docking minimization will not improve the results for poses that do not fit
or are already poorly aligned to the receptor.
For HTVS and SP docking, the minimized poses replace the original poses. For XP docking,
the pose is kept if the GlideScore is no higher than that of the original pose by less than a
certain threshold, otherwise the original pose is kept. This threshold is specified in the
Threshold for rejecting minimized pose text box.
The time taken for post-docking minimization is less than 1% of the total docking time for SP
and XP docking, and can be around 10% for HTVS.
In addition to minimizing, you can choose to calculate and apply a strain correction when
doing the final scoring, by selecting Apply strain correction terms. These terms are evaluated by
optimizing each ligand pose as a free ligand, first with constraints on all torsions, then without
these constraints. The difference is used to compute a penalty term for unreasonably high
strain: the strain correction is only added if it is above a threshold, and the excess strain above
this threshold is scaled before adding it to the GlideScore. You can set the threshold and the
scaling factor in the Output - Advanced Settings dialog box.
5.8.3
Per-Residue Interaction Scores
To examine interactions between a ligand and individual, nearby receptor residues, you can
write out per-residue interaction scores for selected residues or for residues within a chosen
distance of the grid center. The Coulomb, van der Waals, and hydrogen bonding scores, the
sum of these scores (interaction energy, Eint), and the minimum distances are calculated
between the ligand and the specified residues. These values are written as properties for each
ligand to the output file, as well as to the log file.
To enable writing of the scores, select Write per-residue interaction scores, then choose an
option to specify the residues:
• For residues within N Å of grid center—Write scores for complete residues that have any
atom within the specified distance of the grid center.
• Pick residues to include—Pick residues for which the scores are written. Click Specify
Residues and use the picking controls in the dialog box that opens to pick residues.
Per-residue interactions can be visualized in the Workspace when viewing poses. See
Section 6.1 on page 79 for more information.
74
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
5.8.4
Advanced Settings
In the Output - Advanced Settings dialog box, you can set options to screen out poses that
either have too high an energy or are too similar to other poses, and change the parameters
associated with strain correction. The thresholds for rejection of poses are set into two sections,
Filter and Clustering; the strain correction thresholds are set in the Strain Correction section.
The Filter section contains a single option, Reject poses with Coulomb-vdW energy greater
than x kcal/mol. If the pose has a coulomb-vdW score greater (more positive) than this value,
the pose is rejected. The default value is 0.0 kcal/mol. This means that poses that interact favorably with the protein site, however weakly, are retained, whereas poses that interact unfavorably are rejected. Change x to a negative value to reject poses with weakly favorable
interactions, or to a positive value to keep poses with mildly unfavorable interactions.
Figure 5.13. The Output - Advanced Settings dialog box.
The clustering options determine which ligand poses are sufficiently alike to be considered
duplicates. To determine whether a pose is a duplicate, Glide uses a sophisticated algorithm for
recognizing “number reorderings” that need to be taken into account—for example, to accommodate trivial variations such as a 180º rotation of a phenyl group. Ligand poses are compared
to those previously selected for inclusion in the reported output, and are discarded as duplicates
if either of the following criteria are met. This ensures that reported poses are conformationally
distinct. The criteria are:
Glide 6.7 User Manual
75
Chapter 5: Ligand Docking
• RMS deviation is less than d Å
For a pose to be considered a duplicate of one already scored the heavy-atom (nonhydrogen) RMS deviation must be less than r Å. The default RMS deviation threshold for r is
0.5 Å; choose any value greater than 0.0 Å.
• Maximum atomic displacement is less than d Å
For a pose to be considered a duplicate the maximum atomic displacement must be less
than d Å. The default is 1.3 Å; choose any value greater than 0.0 Å.
The use of symmetry is somewhat time-consuming, so there is a limit of 100 on the number of
poses that are compared by this method.
In addition, the torsions of hydroxyl and thiol can be checked independently, by selecting
Include hydroxyl/thiol torsions. If the difference in all such torsional angles is less than the value
given in the text box, and one of the RMSD or MAD criteria is also met, the pose is considered
a duplicate.
The GlideScore is corrected for excessive strain of the ligand by adding to it a fraction of the
strain energy that exceeds a threshold value. This correction is made only if you have selected
Apply strain correction terms in the Output tab. The threshold value is specified in the
Threshold for strain correction text box, and the fraction of the excess strain energy, expressed
as a decimal number, is specified in the Scaling factor for excess strain energy text box. The
default value of the threshold is 4.0 and of the scaling factor is 0.25.
You can also set an option to keep the zip archive, jobname_subjob_poses.zip, which
contains the pose files from all the subjobs of a distributed docking job. The archive is copied
back to your working directory. This archive may be useful if you limit the number of poses to
report for the job too severely and want to examine other poses without doing the entire
docking run again. You can run glide_merge using the contents of this archive with the -n
option to set a new number of poses to report. If disk space is an issue, deselect this option.
5.9
Docking Output Properties
In addition to the docked poses, HTVS and SP docking runs generate a number of Maestro
properties, including terms in the scoring function; various indexes; and ligand efficiency
metrics, which allow discrimination on the basis of size for ligands that have similar docking
score values. These properties are listed in Table 5.3. See page 10 for a description of the
GlideScore terms and energy components.
76
Schrödinger Software Release 2015-2
Chapter 5: Ligand Docking
Table 5.3. Maestro properties generated by HTVS and SP docking runs.
Property
Description
glide lignum
Ligand number
docking score
Docking score, including all additional terms.
glide gscore
GlideScore
glide lipo
Lipophilic contact plus phobic attractive term in the GlideScore.
glide hbond
Hydrogen-bonding term in the GlideScore.
glide metal
Metal-binding term in the GlideScore.
glide rewards
Various reward or penalty terms
glide intrahb reward
Rewards for ligand intramolecular hydrogen bonds. Also included in
glide rewards.
glide evdw
Van der Waals energy.
glide ecoul
Coulomb energy.
glide erotb
Penalty for freezing rotatable bonds in the GlideScore.
glide esite
Term in the GlideScore for polar interactions in the active site.
glide emodel
Model energy, Emodel.
glide energy
Modified Coulomb-van der Waals interaction energy
glide einternal
Internal torsional energy
ligand efficiency
(docking score) / (number of heavy atoms)
ligand efficiency sa
(docking score) / (number of heavy atoms)2/3. This efficiency metric
approximates the effect of surface area.
ligand efficiency ln
(docking score) / (1 + ln(number of heavy atoms))
glide confnum
Conformer index for the pose
glide posenum
Pose index of the pose
glide RMSD to input
RMSD between the docked poses and the input geometries. Only present if Compute RMSD to input ligand geometries was selected in the
Output tab.
conssat constraint
Boolean reporting whether the constraint labeled constraint was satisfied. The label includes the name specified for the constraint, and if multiple constraint groups were used, includes a suffix (Group n) to indicate
the group in which the constraint was satisfied.
Glide 6.7 User Manual
77
78
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 6
Chapter 6:
Visualizing Docking Results
Several tools are provided to help visualize the results of docking runs. The View Poses facility
in the Project Table panel enables you to display the ligand poses with the receptor in the
Workspace, along with hydrogen bonds, bad and ugly contacts, and per-residue interaction
information. This facility is described in the next sections.
For Glide SP and XP docking runs, you can visualize the contributions to the XP docking
score, provided that descriptor information was requested in the docking run. Visualization in
the Workspace is done using the Glide XP Visualizer panel, which is described in Section 6.2.
6.1
Viewing Poses
Maestro has a special facility for viewing poses from a pose viewer file (_pv.mae[gz]) or an
ensemble pose viewer file (_epv.mae[gz]). This facility allows you to step through poses
with the receptor fixed in the Workspace. You can also define and display hydrogen bonds,
contacts, and per-residue interactions, and count hydrogen bonds and contacts. If you have
flexible hydroxyl or thiol groups (“rotatable groups”) in the receptor, these are shown in the
correct orientation for each pose, and you can export the receptor and poses for given orientations of the rotatable groups. If you have multiple receptors from an ensemble pose viewer file,
the best receptor for each ligand is shown as you step through the ligands.
These capabilities are available in the Pose Viewer panel, which you can open by choosing
Entry → View Poses in the Project Table panel, or by choosing Tasks → Docking → Pose
Viewer or Applications → Glide → Pose Viewer in the main window. You can also right-click
on an entry group in the Entry List panel or the Project Table panel and choose View Poses to
select the group and open the Pose Viewer panel.
6.1.1
Setting Up Pose Viewing
To use the pose-viewing facility to step through poses, you must select a single entry group in
the Project Table. The group must contain the receptors at the beginning of the group, followed
by the ligands, and the entries in the group must have the appropriate properties. This is the
normal situation when you import a pose viewer file or ensemble pose viewer file into the
project. Ensemble pose-viewer files have multiple receptors at the beginning of the file, and a
property that associates each ligand pose with its receptor. You can create an ensemble poseviewer file from separate pose-viewer files with the glide_merge utility, using the -epv
option. (Hereafter, both kinds of files are referred to as pose viewer files.)
Glide 6.7 User Manual
79
Chapter 6: Visualizing Docking Results
Figure 6.1. The Pose Viewer panel.
To set up pose viewing for a set of poses in the Project Table:
1. Select the entry group in the Project Table that contains the poses.
The receptor or receptors must be at the beginning of the group, followed by the ligands.
This is the normal situation when you import a pose viewer file into the project.
2. Choose Entry → View Poses.
The Pose Viewer panel opens.
3. Click Set Up.
The first ligand entry is included in the Workspace, the receptor is fixed in the Workspace, the and the view zooms to the ligand.
To import poses and set up pose viewing from the Pose Viewer panel:
1. Open the Pose Viewer panel.
2. Click Import File.
3. Navigate to and import the pose-viewer file.
Setup is performed automatically.
80
Schrödinger Software Release 2015-2
Chapter 6: Visualizing Docking Results
To import poses in the Import dialog box and set up pose viewing:
1. Open the Preferences panel (Maestro → Preferences).
2. Under Project – Importing structures, ensure that For pose-viewer files, turn on pose viewing is selected.
Once you have done this, you can skip the steps above the next time.
3. Open the Import dialog box, for example by clicking the Import toolbar button.
4. Navigate to and import the pose-viewer file.
When the file is imported into the project, the Pose Viewer panel opens and setup is performed automatically.
6.1.2
Stepping Through Poses
You can step through the poses using the RIGHT ARROW and LEFT ARROW keys, or the
ePlayer controls on the ePlayer toolbar. For example, you can use the ePlayer to play through
the poses automatically, and perform actions on the poses with a script. For more information
on the ePlayer, see Section 9.7 of the Maestro User Manual.
If your docking run used Epik to expand tautomeric or ionization states in your input ligands,
or you generated conformers outside Glide (e.g. with ConfGen or MacroModel), or you chose
to write out more than one pose per ligand, you might want to step through the top poses for
each ligand rather than all poses. Maestro provides some ways of selecting the top ligands if
the ligands are identified by the value of a property, such as the title. Here we assume that it is
the Title property that identifies the ligands. If the ligands are already sorted by the GlideScore
or the docking score, you can use these two simple steps:
1. Select the entry group containing the ligands.
2. Right-click on the Title column and choose Deselect Duplicate Values.
The first ligand for each value of the property is selected, and the entries for all other occurrences of the property are deselected. You can now step through the ligands with the left and
right arrow keys.
If the ligands are not sorted by the desired property, you can use another approach, which
selects the ligands after sorting:
1. Select the entry group containing the ligands.
Glide 6.7 User Manual
81
Chapter 6: Visualizing Docking Results
2. Click the Sort button on the Project Table toolbar.
The Sort Project Table panel opens.
3. Choose the property to sort by from the list on the left, e.g. docking score or glide gscore.
4. Make sure Sort entries is selected.
5. Click the arrow button under Sort entries, to add the property to the sort list, and ensure
that the sort direction is correct. (If not, click the Direction cell to change it).
6. On the lower right, select After sorting, select top.
7. Select Unique entries in each group, by property.
8. Click the button with the table icon (to the right of the text box) and choose the property
that distinguishes different ligands (e.g. Title)
9. Click Sort Selected.
The Sort Project Table panel closes.
The entries in the group should now be sorted and the top entry for each ligand should be
selected. You can now step through the entries with the left and right arrow keys.
If you want to compare the results of two different docking runs by including one ligand from
each run in the Workspace and stepping through the poses for each run simultaneously, you can
do so as follows.
1. Select all the poses for all the ligands of interest.
If you want the ligands for each run to correspond (same ligand in each run), you will
have to ensure that you select the corresponding ligands and place them in the desired
order—for example, sort them by title.
2. If you want to view the receptor, fix it in the Workspace (right-click on it and choose Fix).
3. Include the first ligand from each run in the Workspace (click the In column for the first,
control-click the In column for the rest).
4. Step through the poses with the RIGHT ARROW and LEFT ARROW keys.
Each time you press one of these keys, each entry in the Workspace is replaced by the
next (or previous) selected entry in the Project Table.
82
Schrödinger Software Release 2015-2
Chapter 6: Visualizing Docking Results
You can also set up the comparison using the pose-viewing mode, which is merely a shortcut
for doing the required setup. The instructions below demonstrate how to do this for two sets of
poses, which are in two entry groups.
1. Set up pose-viewing mode for the first entry group (select it and choose Entry → View
Poses → Set Up).
2. Select the second entry group (control-click) and include the first pose in this entry group
(control-click the In column).
If you want the ligands for each group to correspond (same ligand in each group), you
will have to ensure that you select the ligands that correspond to the first group and place
them in the desired order—for example, sort them by title in both groups.
3. Step through the poses with the RIGHT ARROW and LEFT ARROW keys.
6.1.3
Displaying Nonbonded and Per-Residue Interactions
In addition to displaying the poses, you can display nonbonded interactions, such as hydrogen
bonds, halogen bonds, contacts, pi-pi interactions and pi-cation interactions. You can also
display per-residue interactions if they have been calculated. To do so, choose the appropriate
options in the Pose Viewer panel. The nonbonded interactions between the ligand and the
receptor are turned on for each ligand in the entry group.
If you want to set the parameters that define H-bonds, halogen bonds, and contacts or change
the appearance of the interaction markers, you can do so under Non-bonded interactions in the
Preferences panel (Maestro → Preferences).
Display of per-residue interactions can only be done if you selected Write per-residue interaction scores in the Output tab of the Ligand Docking panel. The display is turned on by selecting
Display in the Per-Residue Interactions section of the Pose Viewer panel. The per-residue interactions are represented by coloring the residues by the value of the interaction, using a color
ramp with white at zero.
To select the residues that are colored and which interactions are visualized, click Define in the
Per-Residue Interactions section of the Pose Viewer panel, and make settings in the PerResidue Interactions panel. In this panel you can select the interaction type, and pick residues
for which the interaction is displayed.
Glide 6.7 User Manual
83
Chapter 6: Visualizing Docking Results
Figure 6.2. The Per-Residue Interactions panel.
6.1.4
Exporting Poses for Rotatable Groups
If you defined a set of receptor rotatable groups (hydroxyls in Ser, Thr, and Tyr residues) in the
Rotatable Groups tab of the Receptor Grid Generation panel (see Section 4.5 on page 41), you
can export the receptor with different orientations of these groups. The coordinates of the
receptor in the pose viewer file are written for the original orientation, but the receptor includes
information on the alternative orientations. This option allows you to write a file with the coordinates for alternative orientations, which you can then use in other applications. (Note that the
receptor is locked in the Project Table, i.e. made read-only, as well as fixed in the Workspace.)
To do so, first include one of the poses that has this orientation in the Workspace. Next, in the
Pose Viewer panel, click Export Rotatable Groups. In the Export Rotatable Groups dialog box
you can choose to export just the receptor, the receptor and the included pose, or the receptor
and all poses for which the receptor has that particular orientation of the rotatable groups.
Figure 6.3. The Export Rotatable Groups dialog box.
84
Schrödinger Software Release 2015-2
Chapter 6: Visualizing Docking Results
6.2
The Glide XP Visualizer
The Glide XP Visualizer panel provides a way to visualize and analyze the results of a Glide XP
docking run. The main functions of this panel are:
• To display the Glide XP results from a pose viewer file (jobname_pv.mae) in a table of
XP terms for each ligand.
• To provide 3D visualizations for XP terms. Information for these visualizations is read
from the pose viewer file. The descriptor file (jobname.xpdes), which is also generated
by Glide XP, can be used instead; it must be in the same directory as the pose viewer file.
• To allow selective evaluation of ligands (and groups of ligands) within the table. This
helps you analyze ligands separately during the screening process.
Before you can use the Glide XP Visualizer, you must generate the descriptor information.
This information is not included in a normal XP run. To generate it, select Write XP descriptor
information in the Settings tab of the Ligand Docking panel. You should also select Write pose
viewer file in the Output tab of the Ligand Docking panel to write the required pose viewer file.
Note:
Generation of XP descriptor information requires a separate license.
To open the Glide XP Visualizer panel, choose XP Visualizer from the Glide submenu of the
Applications menu.
When the panel opens, click Open to select the pose viewer file. If you want to read the
descriptor file instead, select .xpdes before clicking Open. The descriptor file must have the
same base name as the corresponding pose viewer file. The file name and receptor name are
displayed in the File and Receptor text boxes.
After the data has been loaded, the Select Activity Property dialog box opens so that you can
select an activity property to display in the table. Choose a property from the list and click OK
if you want to display the activities in the ligands table, or click Cancel if you do not want to
display an activity property. If you do not choose a property, the Activity column is present, but
shows ‘--’ for the activity. You can change an activity value by clicking in the table cell. When
you do so, the Edit Activity dialog box opens, and you can enter a new value.
Glide 6.7 User Manual
85
Chapter 6: Visualizing Docking Results
Figure 6.4. The Glide XP Visualizer panel.
6.2.1
The Ligands Table
When you open a file, the ligands are loaded into the table, with the name of the ligand in the
far left column, along with the XP terms in the rest of the columns. If the same ligand name is
listed twice in the pose viewer file, then a -N is added to the ligand name, where N is an integer
starting with 2 that is incremented for each successive instance of a ligand name. By default the
first ligand is selected.
To sort the rows of the table by the values in a column, click on the column heading. There are
three sort settings: ascending (+), descending(-), and the original pose viewer order. These
settings can be changed by successively clicking on the column heading. One column can be
sorted at a time. The sorting mechanisms work on the ligands shown in the table.
To change the order of the columns in the table, click Set Column Order, which opens the Set
Column Order dialog box. In this dialog box, you can select the columns from a list and move
them up and down in the list. When you click OK, the columns are reordered with the top
column in the list at the left of the table. You can also order the columns by decreasing values
of the variance, which is listed in the list along with the column name.
The background color of the cells can be chosen to reflect the values of the XP terms. To color
the cells, select Conditional Coloring. To set up the coloring scheme, click Edit, and make
changes in the Edit Conditional Coloring dialog box:
86
Schrödinger Software Release 2015-2
Chapter 6: Visualizing Docking Results
Figure 6.5. The Edit Conditional Coloring dialog box.
• To remove coloring for XP Terms, select them in the Colored XP Terms list, and click
Remove. The terms are transferred to the Available XP Terms list, and the color scheme is
no longer displayed. (The color scheme is still associated with the terms, so it can be reinstated).
• To reinstate coloring for XP Terms, select them in the Available XP Terms list, and click
Add. The terms are transferred to the Colored XP Terms list, and the color scheme is displayed.
• To change the color scheme for a term, select it in the Colored XP Terms list, and use the
controls in the lower half of the panel to set up the scheme. You can choose to apply a
color to terms that are below or above a given value, to color between these thresholds
either with a single color or a color ramp, and to color particular values.
Glide 6.7 User Manual
87
Chapter 6: Visualizing Docking Results
Clicking on cells in the table produces two different functions depending on the cell type:
• XP Term Cells—Cells that have a gray border around them indicate XP term values for
which a visualization exists. By clicking on one of these cells, the cell is highlighted and
the visualization is displayed in the Workspace. These visualizations depend on the XP
term and are described below.
• Ligand Name Cells—You can select multiple ligands by clicking in the Name column,
using shift-click and control-click. If Selected ligands is selected in the Display section,
then Maestro shows the selected ligands and the view changes each time the selection is
changed. There is a limit on the number of ligands you can display. If you select more
than 100 ligands, a warning is posted, and Selected ligands is turned off.
You can step through the ligands in the table using the LEFT ARROW and RIGHT ARROW
keys. If a cell is selected, the arrow keys move up or down the column for that cell, displaying
the visualization if a cell has an associated visualization.
The table columns can be widened so that the entire heading is displayed by deselecting
Narrow columns. The column descriptions are displayed in tooltips. If you want to change the
column order, click Set Column Order and use the tools in the Set Column Order dialog box to
move the columns in the list. You can select multiple columns and move them to the top or
bottom of the list, or move them up or down in the list.
In addition to selecting ligands in the table, you can limit what is shown in the table by using
the Show buttons. The Selected Only button shows only the selected ligands in the table, and
the sorting mechanisms now work exclusively on the ligands shown in the table. The All button
shows all ligands in the table.
The Ligands table also lists a set of fragments that maximize the values of the XP descriptors,
if these have been docked along with the ligands. These fragments are a special set of fragments that are docked using the Virtual Screening Workflow, by selecting Compute maximum
values by docking fragments in the Docking Options tab. The fragments are displayed in the
Workspace with purple carbons, to distinguish them from the ligands. You can display the
scores for the ligands relative to the maximum value for each descriptor by selecting Relative
Scores. Positive values of the relative score mean that the ligand score does not achieve the
maximum as obtained from the fragment set, and indicates room for improvement.
6.2.2
Controlling the Display
The normal Workspace controls to rotate, translate, and zoom in on the structure can be used.
The Glide XP Visualizer panel provides some additional control over the display.
To change the view so that the ligands that are displayed occupy the entire Workspace, click
the Fit to Ligand button.
88
Schrödinger Software Release 2015-2
Chapter 6: Visualizing Docking Results
To control what is displayed in the Workspace, you can use the Display options:
• Receptor—Display or undisplay the receptor. By default, the receptor is displayed. Only
the residues within 5 Å of the ligand are displayed.
• Selected ligands—Display the ligands that are selected in the table. If this option is
selected, the Workspace is updated whenever the selection of ligands is changed.
Note:
If multiple ligands are selected, changing the view might take a considerable
amount of time. Unless it is necessary to view many ligands at once, it is recommended that this option is used infrequently.
• XP Waters—Waters are shown as small red spheres in the visualizations and can be displayed or hidden using this option. By default, waters are not displayed.
• Hydrophobic/philic map—Hydrophobic and hydrophilic surfaces can be shown in the
visualizations by selecting this option. The first time this option is selected, the HPPMap
job is run, and the button flashes until the job finishes. The surfaces are stored for subsequent use. The job takes 1–2 minutes.
6.2.3
Exporting Data and Structures
You can export the information shown in the table to a CSV file. To do so, click Export Data. A
file chooser opens, in which you can navigate to a location and name and save the file. If you
want to export data for a selection of ligands, select them in the table and click Selected Only
before clicking Export Data.
You can also export structures from the table to a Maestro file, along with the receptor structure. The receptor is the first structure, followed by the ligands, just as in a pose-viewer file. All
selected ligands are exported, or if there is no selection, all ligands that are shown in the table
are exported.
6.2.4
XP Terms and Their Visualizations
The XP terms and a description of their visualizations is given in Table 6.1. In the visualizations, hydrogen atoms are not generally displayed, and the ligand carbon atoms are colored
green.
In addition to the XP terms, you can display similarities between the ligands in XP descriptor
space, by selecting Similarities. The similarity is only displayed if a single ligand is selected in
the table.
Glide 6.7 User Manual
89
Chapter 6: Visualizing Docking Results
Figure 6.6. Visualization of hydrophobic enclosure terms.
Table 6.1. Glide XP terms and their visualizations
90
XP Term
Description
Visualization
GScore
Total GlideScore; sum of XP
terms
None
LipophilicEvdW
Lipophilic term derived from
hydrophobic grid potential at
the hydrophobic ligand atoms.
None
PhobEn
Hydrophobic enclosure reward
Hydrophobic atoms on the protein that are necessary for recognition of hydrophobic enclosure are displayed in CPK representation in
gray. Hydrophobic atoms on the ligand necessary for hydrophobic enclosure are displayed in
green in ball and stick representation.
PhobEnHB
Reward for hydrophobically
packed H-bond
H-bonds are displayed as pink dotted lines.
Hydrophobic atoms on the protein that enclose
hydrophobic groups on the ligand are displayed
in gray in CPK representation. Hydrophobic
atoms on the ligand are displayed in ball and
stick representation.
Schrödinger Software Release 2015-2
Chapter 6: Visualizing Docking Results
Table 6.1. Glide XP terms and their visualizations (Continued)
XP Term
Description
Visualization
PhobEnPairHB
Reward for hydrophobically
packed correlated H-bonds
H-bond are displayed as pink dotted lines.
Hydrophobic atoms on the protein that enclose
hydrophobic groups on the ligand are displayed
in gray in CPK representation. Hydrophobic
atoms on the ligand are displayed in ball and
stick representation.
HBond
ChemScore H-bond pair term
H-bonds are displayed as yellow dotted lines.
Electro
Electrostatic rewards; includes
Coulomb and metal terms.
Relevant atoms in the ligand are displayed in
ball-and-stick representation.
SiteMap
SiteMap ligand-receptor non-Hbonding polar-hydrophobic
terms
Hydrophobic (orange) and hydrophilic (turquoise) surfaces are displayed, representing the
field of the receptor. These are the same as the
Maestro Hppmap surfaces.
π Cat
Reward for pi-cation interactions
The relevant aromatic groups in the protein are
displayed in CPK representation and in the
ligand in ball and stick representation.
ClBr
Reward for Cl or Br in a hydrophobic environment that pack
against Asp or Glu
None
LowMW
Reward for ligands with low
molecular weight
None
Penalties
Polar atom burial and desolvation penalties, and penalty for
intra-ligand contacts
Atoms are labeled with the penalties. The
labels and their descriptions and the associated
visualizations are as follows:
Charge+value—Charge penalty. Penalty for
putting a ligand charge in a region without
water and with the charge not in a salt bridge.
Ball and stick representation for ligand
atoms.
Water+value—Water-protein and water-ligand
penalties. Penalties for desolvating (via presence of ligand) a polar group on the protein
or ligand and not making an H bond to the
polar group. Ball and stick representation for
ligand and protein atoms
Glide 6.7 User Manual
91
Chapter 6: Visualizing Docking Results
Table 6.1. Glide XP terms and their visualizations (Continued)
XP Term
Description
Visualization
DDpen+value—Donor-donor penalty. Penalty
for non-H-bonded donors on the ligand and
protein too close together and with semi-collinear NH--HN.
Blk+value—Penalty for burial of charged
group on the protein by the ligand with no H
bonds made to the charged group by the
ligand or protein. Ball and stick representation for protein atoms
Polar+value—Penalty for desolvation of a
polar ligand atom in a hydrophobic protein
environment. Similar to water-ligand penalty
but specific to phobic environments. Ball and
stick representation for protein atoms
Contact+value—Penalty for intraligand con-
tacts. Ball and stick representation for ligand
atoms
PhoBl+value—Phobic ligand-protein block-
age. Penalty for putting a phobic group of the
ligand against polar (donor/acceptor) groups
of the protein in a protein region which normally would give a favorable phobic packing
score (PhobEn). Ball and stick representation
for protein atoms
Twisted amide linkages in the ligand are also
displayed in ball-and-stick, but not labeled.
92
HBPenal
Penalty for ligands with large
hydrophobic contacts and low
H-bond scores
None
ExposPenal
Penalty for solvent-exposed
ligand groups; cancels van der
Waals terms.
The molecular surface of the protein around the
ligand is displayed as a semi-transparent surface. Ligand groups that are exposed are rendered in ball-and-stick representation
RotPenal
Rotatable bond penalty
Rotatable bonds in the ligand are displayed in
tube representation.
Schrödinger Software Release 2015-2
Glide User Manual
Chapter 7
Chapter 7:
Running Glide from the Command Line
This chapter contains information on running Glide from the command line. Although you will
usually set up Glide jobs using the controls and settings in the Maestro GUI, you sometimes
might want to submit jobs from the command line for the following reasons:
• The command-line scripts can run all full-featured jobs written using the Glide panels in
Maestro, and also allow you to override specific run-time values that are not accessible
through the Maestro interface.
• Command-line scripts allow you to run Glide jobs when you want.
• Input files and scripts can be modified and jobs can be re-run without reconfiguring and
reloading job settings in Maestro.
Using Maestro is the best way to write Glide input files, even if you intend to run them from
the command line, because the files are syntactically correct. To create the input files needed to
run a job from the command line, set up the job in Maestro, click the panel’s Write button, and
modify the files that are generated.
Glide jobs are ultimately run using the Impact program, but they are started with the glide
command. This command reads a Glide input file, processes it and runs the Impact executable.
You can run both single-processor and distributed Glide docking jobs with this command.
Glide jobs can also be run with the impact and para_glide commands, but these commands
require a DICE (Impact) input file, and para_glide is deprecated. See Appendix A for information on these commands.
Glide also has three command-line utilities, kept in the utilities subdirectory of the installation: glide_sort, glide_merge, and glide_rescore, which are also described in this
chapter.
For information about general-purpose utilities, see the General Utilities manual.
Glide 6.7 User Manual
93
Chapter 7: Running Glide from the Command Line
7.1
Running Jobs
To run jobs on Linux or Mac, open a terminal window and set the SCHRODINGER environment
variable to the Schrödinger software installation directory, as follows:
csh/tcsh:
setenv SCHRODINGER installation-directory
bash/ksh:
export SCHRODINGER=installation-directory
Once you have defined this environment variable, you should use it to define the paths to the
applications and utilities, e.g. $SCHRODINGER/glide. In the text below, the path will not be
given.
On Windows, you can use a Schrödinger Command Prompt (or Power Shell) and sh to run
Linux commands (see Section 1.1 on page 1). When you open this shell, the SCHRODINGER
environment variable is already set and in your path, as is the utilities folder, so you can
run applications and utilities with just the name.
Command-line applications are located in the main Schrödinger installation directory;
command-line utilities are located in the utilities subdirectory. For usage summary information on any application or utility, use the -h (help) option:
command -h
utilityname -h
Schrödinger applications and utilities run under Schrödinger’s Job Control facility. To run jobs
on remote hosts, you must set up a hosts file. For information on this file and other aspects of
configuring hosts and Job Control, see Chapter 7 of the Installation Guide. You can force
Impact to run outside Job Control, but you are then responsible for managing all the files.
Jobs are automatically run in the background. You need not add an & at the end of the
commands to have them run and immediately display the command prompt. The -WAIT option
forces the shell to wait until the job is finished, so you can embed such commands in other
scripts.
Once your jobs are launched, you can monitor their progress using the Monitor panel in
Maestro. The command jobcontrol can also be used. It has many options, but the two most
useful options are:
jobcontrol -list
which shows the status of all your jobs, and:
jobcontrol -kill jobid
to terminate any job and its subjobs, if any exist.
94
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
For a summary of jobcontrol options, use:
jobcontrol -h
For more information, see the Job Control Guide.
7.2
Job Files and Directories
Location of Files and Working Directory
For Glide jobs, Maestro normally writes input files to a subdirectory of the directory from
which you launched Maestro (the Maestro working directory). This subdirectory (folder) is
named with the job name. The output files are also written to this subdirectory. For grid generation, the Glide interface allows you to specify an arbitrary location for the grid zip file.
File Name Conventions
A typical job has one command-script file (jobname.in or jobname.inp), one or more structure files (jobname.mae, jobname.pdb, or jobname.sdf), and after execution, several output
files (e.g., jobname.out for textual data). Docking jobs can also use structures from a Phase
database (.phdb), which must be specified in the input file instead of the ligand files, and must
be present on the host the job is run on (it is not copied).
If a file already has the name of an output file, in many cases the old file is renamed with a
numerical extension (filename.out.01, filename.out.02, and so on) for archival purposes.
The new job’s output is then written to the base name (filename.out). If you do not need the
old files, you can remove them.
Some files, such as jobname.log files, are newly written each time a calculation is run. Likewise, old jobname_pv.suffix files are overwritten. Glide writes intermediate Maestro-format
structure output to jobname_raw.suffix files, which are incremented. Output structure files can
be written in Maestro or SD format, and can be compressed or uncompressed.
Table 7.1 and Table 7.2 contains descriptions of the various files generated during Glide grid
generation and Glide docking. For more information, see the Maestro online help or the Impact
Command Reference Manual.
Glide 6.7 User Manual
95
Chapter 7: Running Glide from the Command Line
Table 7.1. Input and output files for Glide grid generation jobs
File
Description
jobname.in
Glide input file, required as input to the glide command. Glide input files
are formatted plain-text files. Maestro creates Glide input files for job submission. You can create or edit them manually with a text editor.
jobname.inp
Impact input file, required as input to the impact command. Impact input
files are formatted plain-text files written in the Impact input file language,
DICE. You can create or edit them manually with a text editor.
jobname.mae,
jobname.mae.gz,
jobname.maegz
Maestro format file containing the receptor structure, compressed or uncompressed. If written by Maestro, this file is named jobname.maegz.
jobname.zip
Zip file containing the receptor grid.
jobname.log
Log file, which captures standard output and standard error messages in text
form. This file is overwritten during subsequent runs.
jobname.out
Output file, containing output from standard output but not standard error.
Output files are appended with numerical extensions when the input file is
used again. Up to 99 output files are retained.
Table 7.2. Input and output files for Glide docking jobs.
96
File
Description
jobname.in
Glide input file, required as input to the glide command. Glide
input files are formatted plain-text files written in the Glide simplified input format. Maestro creates Glide input files for job submission. You can create or edit them manually with a text editor.
jobname.inp
Impact input file, required as input to the impact command.
Impact input files are formatted plain-text files written in the
Impact input file language, DICE. You can create or edit them manually with a text editor.
jobname.mae[[.]gz]
Maestro format file containing input structures for docking.
jobname.log
Log file. If specified, a .log file captures standard output and standard error messages in text form. This file is overwritten during
subsequent runs.
jobname.out
Output file, containing output from standard output but not standard error. Output files are appended with numerical extensions
when the input file is used again. Up to 99 output files are retained.
jobname.rept
Plain text file containing a table of ranked poses, scores, and score
components.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
Table 7.2. Input and output files for Glide docking jobs. (Continued)
File
Description
jobname_lib.maegz or
jobname_lib.sdfgz
Glide ligand structure output file, in Maestro or SD format, compressed.
jobname_pv.maegz
Glide pose viewer file, in Maestro format, compressed. Contains
the receptor structure, followed by all output ligand poses ranked
by their score.
jobname_raw.maegz or
jobname_raw.sdfgz
Glide’s intermediate structure files, containing unranked ligand
poses.
jobname_subjob_poses.zip
Zip archive containing pose files for subjobs, when the docking job
is distributed. The subjob pose files are removed after they are successfully archived.
7.3
The glide Command
The glide command is used for both receptor grid generation and ligand docking jobs. For
ligand docking, it can run both single processor and distributed docking jobs. The syntax of the
command is as follows:
glide [options] input-file
For a description of the command options, run the command glide -h.
The input file is a Glide input file, whose default extension is .in. This file contains keywordvalue pairs that define the parameters of the calculation. These keywords are listed in the tables
below. Table 7.3 contains keywords for grid generation jobs, and Table 7.4 contains keywords
for ligand docking jobs. The keywords for grid generation and ligand docking are grouped
according to the tabs (and associated Advanced Settings dialog boxes) in the Receptor Grid
Generation and Ligand Docking panels in which they are set. Defaults are available for many of
the keywords, so you do not need to include all the keywords in the input file.
Boolean values in the input file can be represented by True or False, Yes or No, 1 or 0. The
last two pairs are converted internally to True or False. The values are case-insensitive. A
Boolean value is denoted by boolean in the tables below, and the value is represented by True
or False, which you can change as you will.
Glide 6.7 User Manual
97
Chapter 7: Running Glide from the Command Line
Table 7.3. Glide input file keywords for grid generation jobs.
Keyword
Description
GRIDFILE filename
Filename of the grid .zip file, if compressed, or the .grd file.
OUTPUTDIR path
Output directory. Must be a path that is accessible on the host
that runs the job. The default is the job launch directory.
CV_CUTOFF value
Coulomb-van der Waals cutoff.
Receptor keywords
98
RECEP_FILE filename
Receptor file name. Must be a Maestro file (.mae, .maegz,
.mae.gz). If LIGAND_MOLECULE is used, it must contain the
ligand as well, which is removed for the grid generation.
RECEP_VSCALE value
General van der Waals radius scaling factor. Default: 1.0.
RECEP_CCUT value
General van der Waals radius scaling partial charge cutoff.
Default: 0.25.
GLIDE_RECEP_MAESCALE boolean
Read per-atom scale factors from the input structure file if set
to True. Default: False.
GLIDE_RECEP_ASLSCALE boolean
Specify per-atom scale factors by ASL if set to True.
Default: False.
ASLSTRINGS list
Comma-separated list of quoted ASL expressions, one expression for each set of residues to be scaled. Only used if
GLIDE_RECEP_ASLSCALE is set to True.
GLIDERECEPTORSCALECHARGES
list
Comma-separated list of charge scaling factors, one factor for
each set of residues that is defined in the ASLSTRINGS list.
GLIDERECEPTORSCALERADII list
Comma-separated list of van der Waals scaling factors, one
factor for each set of residues that is defined in the
ASLSTRINGS list.
REC_MAECHARGES boolean
Use partial charges from the input structure if set to True.
Default: False.
HBOND_ACCEP_HALO boolean
Accept halogens (neutral or charged, F, Cl, Br, or I) as H-bond
acceptors. Default: False.
HBOND_DONOR_HALO boolean
Accept the halogens (Cl, Br, I, but not F) as potential “Hbond” (noncovalent interaction) donors. Default: False.
HBOND_DONOR_AROMH boolean
Accept aromatic hydrogens as potential H-bond donors.
Default: False.
HBOND_DONOR_AROMH_CHARGE
value
Partial charge cutoff for accepting aromatic hydrogens as
potential H-bond donors. The cutoff is applied to the actual
(signed) charge, not the absolute value. Default: 0.0.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
Table 7.3. Glide input file keywords for grid generation jobs. (Continued)
Keyword
Description
PEPTIDE boolean
Set up grids for peptide docking. Required if peptides are to be
docked.
Site keywords
INNERBOX x,y,z
Comma-separated list of the x, y, and z dimensions of the inner
(ligand centroid bounding) box in angstroms. These values
must be integers. Written to the receptor Maestro file as
r_glide_gridbox_ligqrange, where q is x, y, or z.
ACTXRANGE value
Outer box x dimension. Must be set to the same as for the
OUTERBOX keyword.
ACTYRANGE value
Outer box y dimension. Must be set to the same as for the
OUTERBOX keyword.
ACTZRANGE value
Outer box z dimension. Must be set to the same as for the
OUTERBOX keyword.
OUTERBOX x,y,z
Comma-separated list of the x, y, and z dimensions of the outer
(grid) box in angstroms. These values can be floating-point
numbers. Written to the receptor Maestro file as
r_glide_gridbox_qrange, where q is x, y, or z.
GRID_CENTER x,y,z
Comma-separated list of the x, y, and z coordinates of the center of the grid. These values can be floating-point numbers.
Written to the receptor Maestro file as
r_glide_gridbox_qcent, where q is x, y, or z.
GRID_CENTER_ASL asl
Specify the grid center with the ASL expression asl. The center is placed at the centroid of the atoms defined in the ASL
expression.
LIGAND_MOLECULE number
Specify the molecule number of the ligand that defines the
grid center.The center is placed at the centroid of the ligand
atoms. If this keyword is used, the receptor file must contain
the ligand, which is removed for the grid generation.
Constraints keywords
GLIDECONS boolean
Define constraints if set to True. Default: False.
GLIDE_NTOTALCONS ncons
Total number of constraints defined (Positional/NOE+H-bond/
Metal+Hydrophobic+Metal Coordination).
GLIDE_NUMPHOBCONS nhpc
Number of hydrophobic constraints defined.
GLIDE_NUMPOSITCONS npc
Number of positional constraints defined.
Glide 6.7 User Manual
99
Chapter 7: Running Glide from the Command Line
Table 7.3. Glide input file keywords for grid generation jobs. (Continued)
Keyword
Description
GLIDECONSATOMS list
Comma-separated list of atom numbers that define H-bond/
Metal constraints.
GLIDECONSNAMES list
Comma-separated list of constraint names.
GLIDECONSUSESYMATOMS list
Include symmetry-related atoms when applying constraints.
List of Booleans for each H-bond or metal constraint, which
matches the list given for GLIDECONSATOMS. Default: True
for all H-bond and metal constraints.
GLIDE_NUMMETCOORDCONS value
Number of metal-coordination constraints (number of metals
selected for coordination constraints).
GLIDE_NUMMETCOORDSITES list
List of the number of available coordination sites for each
metal coordination constraint.
GLIDE_CONS_XMETCOORD list
GLIDE_CONS_YMETCOORD list
GLIDE_CONS_ZMETCOORD list
Comma-separated list of coordinates of the metal coordination
sites. The coordinates of the sites for each metal center should
be contiguous, and they should be in the order implied by the
list defined by GLIDE_NUMMETCOORDSITES.
GLIDE_CONS_RMETCOORD list
Comma-separated list of radii of the metal coordination
spheres, which represent the maximum allowed distance of a
ligand atom from the coordination site.
Rotatable groups keywords
USEFLEXMAE boolean
Include flexible groups in the grid generation, as specified in
the Maestro file for the receptor. Only one of USEFLEXMAE
and USEFLEXASL can have the value True.
USEFLEXASL boolean
Include flexible groups in the grid generation, as specified by
the FLEXASL keyword. Only one of USEFLEXMAE and
USEFLEXASL can have the value True.
FLEXASL asl
ASL expression that specifies the hydrogens on the flexible
groups. The expression must specify only the required
hydroxyl and thiol hydrogens.
Excluded volumes keywords
100
GLIDE_NUMEXVOL value
Number of excluded volumes. For each excluded volume, the
coordinates of the center and the radius must be defined, at a
minimum.
GLIDE_XEXVOL list
GLIDE_YEXVOL list
GLIDE_ZEXVOL list
Comma-separated list of coordinates of the centers of the
excluded volume spheres.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
Table 7.3. Glide input file keywords for grid generation jobs. (Continued)
Keyword
Description
GLIDE_REXVOL list
Comma-separated list of radii of the excluded volume spheres.
GLIDEXVOLNAMES list
Comma-separated list of names of the excluded volumes.
GLIDE_REXVOLIN list
Comma-separated list of the inner radii of the excluded volume spheres (the radii at which the penalties have their full
values). Default is 90% of the sphere radius.
Table 7.4. Glide input file keywords for ligand docking jobs.
Keyword
Description
GRIDFILE filename
File name of the grid .zip file, if compressed, or the .grd file.
Settings keywords
DOCKING_METHOD
{confgen|rigid|
mininplace|inplace}
Docking method:
confgen—Dock flexibly.
rigid—Dock rigidly.
mininplace—Refine (do not dock)
inplace—Score in place (do not dock)
PRECISION {XP|SP|HTVS}
Docking precision mode. Set in the Settings tab.
CANONICALIZE boolean
Canonicalize the input structure by discarding the coordinates
and regenerating the structure from the connectivity and stereochemistry. Takes about 1 sec per ligand. Default: False.
SAMPLE_N_INVERSIONS
boolean
Sample nitrogen inversions if set to True and
DOCKING_METHOD is set to confgen. Default: True.
SAMPLE_RINGS boolean
Sample rings if set to True and DOCKING_METHOD is set to
confgen. Default: True.
INCLUDE_INPUT_RINGS
boolean
Include the input structure along with any generated ring conformations. Only applies if SAMPLE_RINGS is set to True (otherwise the input structure is the only structure). Default: False.
RINGONFLY boolean
Generate templates dynamically for unrecognized rings of up to
14 atoms. Only available with OPLS_2005. Default: False.
AMIDE_MODE {penal|free|
fixed|trans}
Amide bond sampling mode. Only set when DOCKING_METHOD
is set to confgen.
penal—penalize nonplanar conformation
free—vary conformation
fixed—retain original conformation
trans—allow trans conformation only
gen[eralized]—use generalized torsion controls defined in
torcontrol.txt.
Glide 6.7 User Manual
101
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
102
Keyword
Description
AMIDE_TRANSTOL angle
Maximum angle deviation in degrees from 180° for an amide to
be considered trans. Default: 20.
EPIK_PENALTIES boolean
Apply penalties for ionization or tautomeric states calculated by
Epik. Default: False.
SKIP_EPIK_METAL_ONLY
boolean
Skip Epik-generated states of ligands that are designed for binding to metals. This option is useful if the receptor has a metal but
the ligand does not bind to it. These states are skipped by default
if the receptor does not have a metal.
EXPANDED_SAMPLING boolean
Expand the sampling by bypassing the elimination of poses in
the rough scoring stage. Useful for fragment docking.
Default: False.
REWARD_INTRA_HBONDS boolean
Reward intramolecular ligand hydrogen bonds by adding a contribution for each intramolecular hydrogen bond to the
GlideScore, and a contribution to Emodel. Default: False.
HBOND_DONOR_AROMH boolean
Accept aromatic hydrogens as potential H-bond donors.
Default: False.
HBOND_DONOR_AROMH_CHARGE
value
Partial charge cutoff for accepting aromatic hydrogens as potential H-bond donors. The cutoff is applied to the actual (signed)
charge, not the absolute value. Default: 0.0.
HBOND_ACCEP_HALO boolean
Accept halogens (neutral or charged, F, Cl, Br, or I) as H-bond
acceptors. Default: False.
HBOND_DONOR_HALO boolean
Accept the halogens (Cl, Br, I, but not F) as potential “H-bond”
(noncovalent interaction) donors. Default: False.
GLIDEUSEALLEXVOL boolean
Apply all excluded volumes that are included with the grid.
GLIDE_EXVOL_PENAL_STRENGTH
value
Set the strength of the penalty for occupying an excluded volume. Allowed values: large, medium, small. These values are
translated into a numeric value for each volume. Default: large.
GLIDE_NUMUSEXVOL value
Number of excluded volumes to use in docking. Must not be
greater than the total number, specified by GLIDE_NUMEXVOL.
for grid generation.
GLIDEUSEXVOLNAMES list
Comma-separated list of names of the excluded volumes to use
in docking.
GLIDE_EXVOL_PENAL_NUM list
Comma-separated list of maximum penalty values for each
excluded volume used in docking.
WRITE_XP_DESC boolean
Write XP descriptor information if set to True. Default: False.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
Keyword
Description
FLEXTORS boolean
Set to False for rigid, refineinput and inplace jobs.
MAXKEEP nposes
Number of poses per ligand to keep in initial phase of docking.
Default: 5000.
SCORING_CUTOFF cutoff
Scoring window for keeping initial poses. Default: 100.0.
MAXREF nposes
Number of poses to keep per ligand for energy minimization.
Maestro sets this number to 800 for XP. Default: 400.
GLIDE_DIELCO constant
Distance-dependent dielectric constant. Default: 2.0.
FORCEPLANAR boolean
Scale force-field parameters for planar (sp2) systems such as
rings to strongly penalize nonplanar conformations.
Default: False.
MAX_ITERATIONS niter
Maximum number of conjugate gradient steps. Default:100.
PEPTIDE boolean
Set parameters for improving docking results for polypeptides.
The following settings are made if the keywords are not already
set:
MAXKEEP
100000
MAXREF
1000
POSTDOCK_NPOSE
100
Other keywords are set internally: the maximum number of conformers is increased by a factor of 10, and the number of diameter directions is increased by about a factor of 3.
It is recommended to follow docking with Prime MM-GBSA
postprocessing, for minimization of the peptide and for scoring.
Not available for HTVS docking precision.
Ligands keywords
GLIDECONSUSEMET boolean
Use individual metal and ligand radii to determine the cutoff for
metal-ligand constraints (rmetal+rligand+0.4), rather than a single
global value of 2.4 Å. Default: False.
GLIDESCORUSEMET boolean
Use individual metal radii for the cutoff in the scoring function,
rather than a single ramp from 2.2 to 2.6 Å. Default: False.
LIGANDFILE filenames
Comma-separated list of ligand file names. File names that
include commas can be quoted; leading and trailing spaces are
ignored. Files can be in any of Maestro, SD, MOL2, or PDB format.If LIGAND_START and LIGAND_END are used, they refer to
the ligand indices as if the input files were concatenated in the
order specified.
LIGAND_START firstlig
First ligand from the ligand file to be docked. Default: 1.
Glide 6.7 User Manual
103
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
Keyword
Description
LIGAND_END lastlig
Last ligand from the ligand file to be docked. Default: last in
file.
PHASE_DB dbpath
Full path to Phase database to use for the ligands. The first conformer of each ligand is extracted from the database for docking.
Use instead of LIGANDFILE, LIGAND_START, LIGAND_END.
PHASE_SUBSET filename
Name of subset file to use in the Phase database. Only the
ligands whose records are listed in the subset file are docked. If
omitted, all ligands in the database are docked.
LIG_MAECHARGES boolean
Take charges from input structures if set to True.
Default: False.
MAXATOMS maxatoms
Maximum number of atoms per ligand.
Default (and maximum): 300.
MAXROTBONDS maxrotbonds
Maximum number of rotatable bonds per ligand.
Default (and maximum): 50
LIG_VSCALE factor
Scaling factor for van der Waals radii scaling. Default: 0.8.
LIG_CCUT cutoff
Partial charge cutoff for van der Waals radii scaling. Default:
0.15.
RINGCONFCUT cutoff
Energy window for retention of ring conformers.
Default: 2.5 kcal mol-1.
Core keywords
104
USE_REF_LIGAND boolean
Use core pattern for comparison or restraint if set to True.
Default: False.
REF_LIGAND_FILE filename
File name for reference ligand. File format can be Maestro, SD,
MOL2 or PDB.
CORE_DEFINITION {all|
smarts|allheavy}
Specify how the core is defined.
all—use all atoms in the reference ligand
allheavy—use all nonhydrogen atoms in the reference ligand
smarts—use the SMARTS pattern defined by the CORE_ATOMS
keyword.
CORE_ATOMS list
List of core atoms to use for the RMSD calculation. Only set
when CORE_DEFINITION is set to smarts.
CORE_SMARTS pattern
Smarts pattern for core. Only set when CORE_DEFINITION is
set to smarts. Set in the Core tab.
CORE_RESTRAIN boolean
Restrict docking to reference position if set to True.
Default: False.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
Keyword
Description
CORE_SNAP boolean
When using core constraints, “snap” the core of the ligand to
exactly match the reference ligand and rebuild the rest of the
ligand with the new core, if set to True. If set to False, use the
filter-and-restrain approach.
Not compatible with CORE_DEFINITION all.
Default: True if CORE_POS_MAX_RMSD < 0.75, False otherwise.
CORE_POS_MAX_RMSD value
Tolerance on the RMSD to use when restricting docking to the
reference position.
CORE_FILTER boolean
Skip ligands that do not match the core pattern if set to True.
Default: False.
Constraints keywords
[FEATURE:n]
Define a feature block. A feature is use to define the ligand
atoms that match a given constraint, and is defined by a set of
SMARTS patterns and rules. The feature block consists of one
or more PATTERN lines. The index n is used to associate the feature with a constraint specified by USE_CONS. The square brackets must be included.
PATTERNi string
Define a ligand matching pattern in a feature block. The index i
must be unique within a block. The string consists of the following components, separated by spaces:
1. SMARTS pattern.
2. Comma-separated list of atoms in the SMARTS pattern that
must be matched by the ligand.
3. Optional keyword include or exclude, for including or
excluding matched atoms. Default is include. exclude can
only be used to limit the scope of an included pattern.
4. For hydrophobic constraints, number of atoms required in the
constraint region. Must be preceded by include.
[CONSTRAINT_GROUP:m]
Define a constraint group. Must be followed by a USE_CONS
keyword and a NREQUIRED_CONS keyword. The square brackets
must be included. All constraints must be defined as part of a
group. The ligand must satisfy the constraints in each group.
USE_CONS list
Apply the specified constraints as part of the constraint group.
The list is a comma-separated list of constraints. Each constraint
is specified in the format title:n, where n must match the index
of a FEATURE block, and title is the label of the constraint in the
.cons file for the grid.
Glide 6.7 User Manual
105
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
Keyword
Description
NREQUIRED_CONS nreq
Number of constraints in the USE_CONS list that are required to
match. The sum of nreq values over all constraint groups must
not exceed 4.
GLIDE_CONS_FINALONLY
boolean
Test constraint satisfaction only after docking if set to True.
Default: False.
HAVEGLIDECONSFEAT boolean
Deprecated, but retained for backward compatibility. A feature
file is already available if set to True. Default: False.
GLIDECONS boolean
Deprecated, but retained for backward compatibility. Use constraints if set to True. Default: False.
GLIDE_CONS_FEAT_FILE
filename
Deprecated, but retained for backward compatibility. Specify the
name of the constraints feature file. Default: jobname.feat.
GLIDEUSECONSFEAT boolean
Deprecated, but retained for backward compatibility. Use the
constraints feature file for the application of constraints if set to
True. If set to False, the constraint information is in the Impact
.inp file.
Torsional constraints keywords
106
[TORCONS:index]
Define a torsional constraint block. If a torsion matches a constraint definition in more than one block, the match in the block
with the highest index. is used. These blocks must be contiguous
and be placed at the end of the input file.
SMARTS "pattern"
Define part of a molecule in which torsional constraints can be
defined by the specified SMARTS pattern. The pattern must be
enclosed in quotes.
ATOMS list
Define torsions to be constrained. The list of atoms that define
the torsions is a comma-separated list of quoted atom number
quartets. The atom numbers are separated by hyphens. For
example:
ATOMS "1-2-3-4","2-3-4-5","4-5-6-7"
The atom numbering is given relative to the SMARTS pattern.
USEVAL list
Specify which torsions should be set to the corresponding values
given in the list for the TORVAL keyword. The list is a commaseparated list of integer values. The default is 0 for all values,
which means that the input value of the torsion will be used.
TORVAL values
Specify the angle values to which the torsions should be set
before constraining them, if the corresponding boolean value in
the USEVAL list is True. The angle values are a comma-separated list.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
Keyword
Description
ALLBONDS boolean
Constrain torsions about all bonds for which a dihedral (defined
by four atoms) is fully contained in the SMARTS pattern, using
the value for each torsion from the input structure. Any ATOMS,
SETVAL, and TORVAL keywords in the constraint block are
ignored.
Output keywords
COMPRESS_POSES
{TRUE|FALSE}
Compress the output pose files. Default: TRUE.
NOSORT {TRUE|FALSE}
Do not sort the poses by score (using glide_sort), but return
the poses as generated in jobname_raw.maegz.
CV_CUTOFF cutoff
Reject poses with Coulomb-van der Waals energy greater than
cutoff kcal/mol. Default: 0.0.
HBOND_CUTOFF cutoff
Reject poses with H-bond score greater than cutoff. Default: 0.0.
METAL_CUTOFF cutoff
Reject poses with metal score greater than cutoff. Default: 10.0.
POSE_OUTTYPE
{poseviewer|ligandlib|
poseviewer_sd|
ligandlib_sd}
Type of output file to produce. Default: poseviewer.
poseviewer—Write pose viewer file (_pv.maegz); includes
receptor as first structure.
poseviewer_sd—Write pose viewer file in SD format
(_pv.sdfgz); receptor stored in V3000 format.
ligandlib—Write ligand pose file (_lib.maegz); does not
contain receptor.
ligandlib_sd—Write ligand pose file in SD format
(_lib.sdfgz).
SD format is not available with flexible groups in the grid or
with XP descriptors.
POSE_RMSD
RMS deviation used in clustering to discard poses, in angstroms.
Default: 0.5.
POSE_DISPLACEMENT
Maximum atomic displacement used in clustering to discard
poses, in angstroms. Default: 1.3.
NMAXRMSSYM
Maximum number of poses to compare taking local symmetry
into account, when detecting duplicates. A negative value means
that there is no limit. Default: 100.
NREPORT maxposes
Maximum number of poses to write out from the docking run.
Default: 0, meaning write out all poses.
OUTPUTDIR path
Directory for output files. Must be a path that is accessible on
the host that runs the job. The default is the job launch directory.
Glide 6.7 User Manual
107
Chapter 7: Running Glide from the Command Line
Table 7.4. Glide input file keywords for ligand docking jobs. (Continued)
108
Keyword
Description
POSES_PER_LIG maxperlig
Maximum number of poses to write per ligand. Default: 1.
POSTDOCK boolean
Use post docking minimization if set to True. Default: True.
POSTDOCK_NPOSE npose
Number of poses to use in post-docking minimization. Maestro
sets this number to 10 for XP. Default: 5.
POSTDOCKSTRAIN boolean
Apply strain correction terms if set to True. Default: False.
WRITE_RES_INTERACTION
boolean
Generate per-residue interaction terms if set to True.
Default: False.
ASL_RES_INTERACTION asl
ASL expression defining residues to be used for per-residue
interactions. Takes precedence over
RADIUS_RES_INTERACTION.
RADIUS_RES_INTERACTION r
Radius for per-residue interaction terms, in angstroms. Only set
if WRITE_RES_INTERACTION is set to True. Default: 12.0
WRITEREPT boolean
Write report text file (.rept for docking and .scor for scorein-place). Default: False.
KEEPSKIPPED boolean
Write a file jobname_skipped.maegz containing the ligands
that were skipped or rejected at any point in the docking process.
Default: False.
KEEP_SUBJOB_POSES boolean
Keep the zip archive, jobname_subjob_poses.zip, and copy
it back to the job launch directory. This archive contains the
pose files from all the subjobs of a distributed docking job.
Default: True.
CALC_INPUT_RMS boolean
Calculate RMSD between each pose and its input structure.
Default: False.
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
7.4
Restarting Glide Docking Jobs
The Job Control facility can automatically restart failed Glide subjobs that were run with the
glide command. The number of times a failed subjob is retried is set by the
SCHRODINGER_MAX_RETRIES environment variable, whose default is 3. This environment
variable is described in Table B.1 of the Job Control Guide.
If you still have failed subjobs when the master job finishes, you can restart the job to run the
subjobs that failed. To do so, use the command that you used to run the job originally with the
-RESTART option added. This restart mechanism only applies to distributed jobs.
7.5
Glide Utilities
This section describes the utilities that are available in the utilities directory of the installation.
7.5.1
glide_sort
Re-ranks Glide poses by custom criteria or combines job outputs into one file. The “best” pose
is defined by the property used for sorting. If you set any of the custom scoring function terms,
sorting is done only on this custom function instead of the standard sorting options. The syntax
of the command is:
glide_sort mode [options] pose-files
The input pose files can be in Maestro or SD format, uncompressed (.mae or .sdf) or
compressed (.maegz,.mae.gz, .sdfgz). The same format must be used for all files, both
input and output. You can combine pose-viewer files (_pv.mae[gz]) into ensemble poseviewer files (_epv.mae[gz]).
At least one of these modes is required:
-o output-file
Write the best-scoring poses to output-file in Maestro or SD format, compressed or
uncompressed. The output format must match the input format, except that
_pv.mae[gz] files can be combined into an _epv.mae[gz] file.
-r report-file
Create a report of the best scores in report-file.
-R
Write a report of the best scores to standard output.
For a description of the command options, run the command glide_sort -h. If multiple
–use_prop[_a|d] options are given, a multi-key sort is used, with the first property specified
as the primary key, the second property as the secondary key, and so on.
Glide 6.7 User Manual
109
Chapter 7: Running Glide from the Command Line
7.5.2
glide_merge
This utility merges results from a list of pose files (such as from glide subjobs) into a single
pose file, sorted by the docking score or a specified property. The pose files must be already
sorted by the docking score or the specified property. The syntax of the command is as follows:
glide_merge mode [options] pose-files
The input and output pose files can be in Maestro or SD format, uncompressed (.mae or .sdf)
or compressed (.maegz,.mae.gz, .sdfgz). The same format must be used for all files, both
input and output. You can combine pose-viewer files (_pv.mae[gz]) into ensemble poseviewer files (_epv.mae[gz]).
At least one of these modes is required:
-o output-file
Write the poses to output-file. If .xpdes files exist for the input pose files, an output
.xpdes file is also created with the same base name.
-r report-file
Create a report of the best scores in report-file.
-R
Write a report of the best scores to standard output.
For a description of the command options, run the command glide_merge -h.
7.5.3
glide_rescore
This utility replaces the docking score properties in Glide pose output files with different
values, so that the glide_sort “best-by-title” option can be used to combine different screens.
See Section 7.5.1. The syntax of the command is as follows:
glide_rescore [options] pv-or-lib-files
For a description of the options, run the command glide_rescore -h. Only one of -rank,
–offset, and -average can be used at a time. If none is specified, -rank is assumed. Using
–top or -every implies -average mode.
7.6
Customizing Glide Calculations
The controls in the Glide panels provide a limited number of options for customizing the
docking calculations. You can make more extensive customizations by adding commands and
keywords to the Impact input file. The syntax of this input file is described in detail in the
Impact Command Reference Manual. You cannot make these customizations with the Glide
input file.
The subsections below describe two customizations that are done by changing resource files.
110
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
7.6.1
Changing the Glide Atom Typing
Glide uses SMARTS patterns to define the atom types (“ptypes”) and hybrid types used in grid
generation and docking. By default, these patterns are read from $SCHRODINGER/impactvversion/data/opls/ptype.def. If you want to change the atom types, you can copy and
edit this file, and then add the appropriate keywords under the PARAMETER subtask in the
Impact (Dice) input file. There are two keywords: ptype, which specifies the file to be used
and must be followed by the file name, and prntptypes, which prints out the ptypes and the
hybrid types.
To customize a grid generation job, an additional parameter statement should be included at the
beginning of the DOCK task:
parameter prntptypes ptype "filename"
For docking jobs, during the first DOCK task when reading the grid files, include:
parameter prntptypes ptype "filename"
The quotes are required to preserve the case of the file name.
Since Glide no longer uses the Dice input file by default, you must convert your Glide input
file to a Dice input file by running glide with the -write_inp option.
7.6.2
Defining Torsional Controls for Planar and Other Groups
The mechanism for controlling rotations about amide bonds in earlier Glide releases has been
generalized to allow control of any class of torsions—for example, around ester linkages. The
torsions and the type of control applied are defined in a resource file, torcontrol.txt,
which is included in the impact-vversion/data directory of the installation. You can override
the default definitions by copying this file to the impact directory in your user resources area
($HOME/.schrodinger on Linux or Mac, %APPDATA%\Schrodinger on Windows), or to
your working directory, and editing the file to provide your own definitions. The file has a
summary of the syntax at the top.
Each torsion control is defined by a keyword for the control type, a unique name, a SMARTS
pattern, and four atom numbers to define the torsion, then values that depend on the control
type. These must all be on the same line, separated by spaces. The control type syntax is:
free
freezeinput
freeze
force
penalize
name
name
name
name
name
smarts
smarts
smarts
smarts
smarts
atoms
atoms
atoms angle
atoms angle tolerance
atoms angle multiplicity penalty barrier exponent
Glide 6.7 User Manual
111
Chapter 7: Running Glide from the Command Line
The various fields in the torsion control are described in Table 7.5.
Table 7.5. Description of fields in the definition of a torsion control.
Field
Description
keyword
Case-insensitive control type. Allowed values:
free—do not constrain the torsion
freezeinput—freeze the torsion at its input value
freeze—freeze the torsion at the specified angle
force—reject poses for which the torsion does not lie within the given tolerance
of the specified angle.
penalize—penalize torsions during scoring that do not adopt the given angle,
using a periodic penalty function with a global minimum (zero penalty) and
other penalized minima.
name
Name of the torsion (case-sensitive). Must be unique within the file, and cannot
contain spaces.
smarts
SMARTS pattern that defines the torsion.
atoms
Indexes of the four atoms in the SMARTS pattern that define the torsion.
angle
Value of the angle (in degrees) to which the torsion is frozen or forced, or at which
the penalty is at its minimum (zero).
tolerance
Tolerance for the angle to which the torsion is forced, in degrees. The tolerance
defines the range of acceptable angles.
multiplicity
Number of maxima (or minima) in the torsional potential. Must be 1, 2, 3, 4, or 6.
If the number is 4 or 6, the potential is assumed to have two global minima separated by 180°, with higher energies for the other minima.
penalty
Penalty value for torsions at the nonglobal minima.
barrier
Maximum value of the torsional potential, which is the barrier between the global
minimum and the next minimum.
exponent
Power to which the torsional potential function is raised (while maintaining the
barrier heights. Useful for flattening the potential around the minima. A value of 2
is recommended to enable some flexibility around the minima.
The order in which the torsions are included in the file matters: the last torsion that is matched
is the one that is used. This means that you should put the more general SMARTS patterns
first, followed by the more specific patterns. You can even include torsions around double
bonds, as the torsion patterns are matched to determine the bonds to be restricted before determining the rotatable bonds.
When looking for matches to a pattern in a molecule, it is possible to find multiple matches—
for example, the pattern CN(=O)cc to constrain the CNcc dihedral angle for an aromatic amide
112
Schrödinger Software Release 2015-2
Chapter 7: Running Glide from the Command Line
matches a molecule that has a benzene ring attached to the amide in two ways, and therefore
the penalty could be applied in two different ways. In such cases, the “favored” orientation is
chosen to be the one that is closest to the input dihedral angle from the molecule.
The freeze and freezeinput controls on the torsions are only applied during docking: they
make the bond non-rotatable. However, post-docking minimization can change these frozen
torsions. The force and penalize controls are applied after post-docking minimization.
An example of a torsion control is given below, for an ester.
penalize
ester
O=C(-O[#6])
1 2 3 4
0.0
2
4.0
9.0
2.0
If a penalty is applied, the name of the torsion control and the penalty value are added as bond
properties (s_glide_torcontrol_name, r_glide_torcontrol_penalty) to the output
structure.
7.7
Docking Log Messages
When a ligand is docked, its fate is reported in messages in the log file. These messages are
described below.
• DOCKING RESULTS FOR LIGAND n—Docking succeeded and at least one pose was
written to the _raw.mae file.
• NO POSES STORED—Nothing was written to the _raw.mae file for this ligand. This usually only happens in constrained jobs, if none of the minimized poses satisfies constraints.
• NO POSES SURVIVED ROUGH-SCORE—There are actually several messages similar to
this (one of which mentions “refinement” as well as “rough-score”). For all of them,
poses have been eliminated somewhere in the rough-score part of the funnel.
• NO GOOD POSES FOUND—This message should occur in conjunction with one of the
two previous messages.
• PROBLEM IN CONFGEN—Problems occured in the conformation generator.
• INVALID ATOMIC NUMBER—An atom in the ligand has an atomic number for which no
atom type can be assigned.
• TOO MANY ROTATABLE BONDS—The user-specified limit on the number of rotatable
bonds was exceeded.
• LIGAND TOO BIG—The ligand has more atoms than either the absolute maximum
allowed (200 atoms) or the maximum set by the user.
Glide 6.7 User Manual
113
Chapter 7: Running Glide from the Command Line
• FAILED INTERCONSTRAINT DIST—More than one constraint was required and the
atoms that could satisfy each constraint individually are in positions that cannot possibly
satisfy the constraints simultaneously.
• TOO MANY ROTAMERS—More conformations were generated for a given rotatable group
than the constraint algorithm can handle. (The limit is 6.)
• HYDROPHOBIC/PERIPHERAL ASSIGNMENT FAILED—Assignment of hydrophobic
regions on the ligand for application of hydrophobic constraints failed.
• ATOMIC COORDINATES OVERLAP—Ligand atoms are too close to each other. This message might indicate poor ligand preparation.
• STRUCTURE INCOMPATIBLE WITH FORCE FIELD—The ligand cannot be used with
the force field selected because the bonding patterns could not be assigned.
• STRUCTURE HAS TOO MANY ATOMS—The ligand has more atoms than Impact can handle. (The limit is 300.)
• INVALID ATOM(S)—The ligand has atoms for which no force field parameters are available.
• ERROR GETTING LIGAND DATA—Glide encountered an error parsing the ligand line of
the (Dice) input file. Usually preceded by a more specific error message.
• ROUGH POSE REFINE FAILED—The preminimization step on the rough poses failed.
This step involves displacing the ligand and sampling rotatable bonds by small amounts
about the generated values.
114
Schrödinger Software Release 2015-2
Glide User Manual
Appendix A
Appendix A:
Old Commands for Running Glide
This appendix provides information on commands for running Glide that are superseded by the
glide command. The impact command can be used to run Glide jobs on a single processor
with a DICE input file. The para_glide command can be used to run Glide jobs on multiple
processors. This command is deprecated and will be removed in a future release. You should
change any scripts that use this command to use the glide command instead.
A.1
The impact Command
The syntax of the command is as follows:
impact [options] [[-i] input-file]
For a description of the command options, run the command impact -h.
If your protein has more than 8000 atoms and you are making grid files (but not if you are just
docking ligands), you must specify a larger executable with the -s option. See Table 6.1 of the
Impact User Manual for more information. If you have a large number of ligands, you can use
para_glide to distribute the docking job over multiple processors
The impact command is also used to run other applications. For more information on the
input file syntax, see the Impact Command Reference Manual.
A.2
The para_glide Command
The para_glide application divides the ligands into batches of ligand structures and submits
these batches to multiple processors for docking. The syntax of the para_glide command is:
para_glide -i input-file [options]
For a description of the command options, run the command para_glide -h. All other
options, such as options for the impact command, are passed to the subjobs.
The para_glide application splits the Glide job specified in the input-file into smaller
subjobs for distributed execution. The ligands from firstlig to lastlig are separated into njobs
equal-sized batches. By default, njobs is set to 1, firstlig is set to 1 and lastlig is set to 0, which
is interpreted as the final ligand in the ligands file.
Glide 6.7 User Manual
115
Appendix A: Old Commands for Running Glide
For example, to submit a job with 50 subjobs to be run on 10 processors, you could use the
following command:
$SCHRODINGER/para_glide -n 50 -HOST mycluster:10 -i myjob.inp
This job would run on the host mycluster using 10 processors.
Each use of para_glide creates two scripts: jobname_report.sh and jobname_status.sh.
The jobname_report.sh script collects the output (poses) from subjobs created by
para_glide, and produces a single pose file and a single report file that summarizes the best
poses in the entire job suite. The subjob results are stored in subdirectories of the working
directory, named jobname_firstlig_lastlig, where firstlig and lastlig are the indexes of the first
and last ligands in the subjob. The jobname_status script can be run at any time to obtain a
summary of the disposition of each job: whether it finished normally, died, was terminated,
stopped, and so on, using the conventional Schrödinger job control terms. (See Chapter 3 of the
Job Control Guide for information about job monitoring and job control.)
The -job option is useful for preparing only a single subjob. A value of 0 is equivalent to not
specifying -job at all, and all subjobs will be prepared. Negative values are not permitted.
By default, a new ligand file is written for each subjob, containing only the ligands for that
subjob. If the -nosplit option is given, all jobs use the same input ligand file, and each job
reads out of it just the ligands pertinent to that job. Splitting the ligand file is likely to be more
efficient if there is a large number of ligands, but requires extra disk space to store the ligand
files.
A.3
Recovering Failed Docking Jobs
For jobs run with impact or para_glide, you can reclaim the work that has been done to
date and rerun jobs for the ligands that have not yet been docked. The results to the point of
failure are stored in a _raw.mae file, which can be used in conjunction with the results of a
new docking run that completes the task to generate the final output file, using glide_sort.
There are several scenarios to consider.
If you ran a single-processor job:
1. Examine the log file to obtain the number of the ligand being docked when docking
failed.
2. Copy the log file to a safe location.
3. Start a para_glide job, specifying the first ligand to be docked as the one that failed.
You can run a para_glide job on a single processor. The results will be in a subdirectory created by para_glide.
116
Schrödinger Software Release 2015-2
Appendix A: Old Commands for Running Glide
If you ran a para_glide job and isolated subjobs failed or didn’t run:
1. Change to the subdirectory for each subjob that failed or didn’t run in turn, and start an
impact job, with the files for the subjob as input. The output is placed in the same subdirectory as the job was started from.
If you ran a para_glide job and a contiguous block of subjobs failed or didn’t run:
1. Determine which subjobs ran from the log files.
2. Run a para_glide job for the block of subjobs that didn’t run.
For subjobs that failed but produced some results, you can rerun them from the point of failure
with para_glide. This procedure creates another subdirectory for the results.
When all the restarted jobs finish, run a glide_sort job from the parent directory, specifying
all the _raw.mae files as input.
Glide 6.7 User Manual
117
118
Schrödinger Software Release 2015-2
Glide User Manual
References
1.
Eldridge, M. D.; Murray, C. W.; Auton., T. R.; Paolini, G. V.; Mee, R. P. Empirical
scoring functions: I. The development of a fast empirical scoring function to estimate
the binding affinity of ligands in receptor complexes. J. Comput.-Aided Mol. Des. 1997,
11, 425–445.
2.
Maggiora, G. M.; Johnson, M. A. Concepts and Applications of Molecular Similarity.
Wiley & Sons: New York, 1990, pp 99–117.
3.
Moreau, G.; Broto, P. The Autocorrelation of a Topological Structure: A New Molecular
Descriptor. Nouv. J. Chim. 1980, 4, 359–360.
4.
Carhart, R. E.; Smith, D. H.; Venkataraghavan, R. Atom Pairs as Molecular Features in
Structure-Activity Studies: Definitions and Applications. J. Chem. Inf. Comput. Sci.
1985, 25, 65–73.
5.
Wagener, M.; Sadowski, J.; Gasteiger, J. Autocorrelation of Molecular Surface Properties for Modeling Corticosteroid Binding Globulin and Cytosolic Ah Receptor Activity
by Neural Networks. J. Am. Chem. Soc. 1995, 117, 7769–7775.
6.
Sheridan, R. P.; Miller, M. D.; Underwood, D. J.; Kearsley, S. K. Chemical Similarity
Using Geometric Atom Pair Descriptors. J. Chem. Inf. Comput. Sci. 1996, 36, 128–136.
7.
Kearsley, S. K.; Sallmack, S.; Fluder, E. M.; Andose, J. D.; Mosley, R. T.; Sheridan, R.
P. Chemical Similarity Using Physicochemical Property Descriptors. J. Chem. Inf.
Comput. Sci. 1996, 36, 118–127.
8.
Pearlman, R. S.; Smith, K. M. Novel Software Tools for Chemical Diversity. Perspectives Drug Discovery Design 1998, 9, 339–353.
9.
Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.;
Repasky, M. P.; Knoll, E. H.; Shaw, D. E.; Shelley, M.; Perry, J. K.; Francis, P.; Shenkin,
P. S. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and
Assessment of Docking Accuracy, J. Med. Chem. 2004, 47, 1739–1749.
10.
Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye, L. L.; Pollard, W. T.;
Banks, J. L. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 2.
Enrichment Factors in Database Screening. J. Med. Chem. 2004, 47, 1750–1759.
11.
Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.; Greenwood, J. R.; Halgren,
T. A.; Sanschagrin, P. C.; Mainz, D. T. Extra Precision Glide: Docking and Scoring
Glide 6.7 User Manual
119
References
Incorporating a Model of Hydrophobic Enclosure for Protein-Ligand Complexes. J.
Med. Chem. 2006, 49, 6177–6196.
120
Schrödinger Software Release 2015-2
Glide User Manual
Getting Help
Information about Schrödinger software is available in two main places:
• The docs folder (directory) of your software installation, which contains HTML and
PDF documentation. Index pages are available in this folder.
• The Schrödinger web site, http://www.schrodinger.com/, In particular, you can use the
Knowledge Base, http://www.schrodinger.com/kb, to find current information on a range
of topics, and the Known Issues page, http://www.schrodinger.com/knownissues, to find
information on software issues.
Finding Information in Maestro
Maestro provides access to nearly all the information available on Schrödinger software.
To get information:
• Pause the pointer over a GUI feature (button, menu item, menu, ...). In the main window,
information is displayed in the Auto-Help text box, which is located at the foot of the
main window, or in a tooltip. In other panels, information is displayed in a tooltip.
If the tooltip does not appear within a second, check that Show tooltips is selected under
General → Appearance in the Preferences panel, which you can open with CTRL+, (,).
Not all features have tooltips.
• Click the Help button in the lower right corner of a panel or press F1, for information
about a panel or the tab that is displayed in a panel. The help topic is displayed in the Help
panel. The button may have text or an icon:
• Choose Help → Online Help or press CTRL+H (H) to open the default help topic.
• When help is displayed in the Help panel, use the navigation links in the help topic or
search the help.
• Choose Help → Documentation Index, to open a page that has links to all the documents.
Click a link to open the document.
Glide 6.7 User Manual
121
Getting Help
• Choose Help → Search Manuals to search the manuals. The search tab in Adobe Reader
opens, and you can search across all the PDF documents. You must have Adobe Reader
installed to use this feature.
For information on:
• Problems and solutions: choose Help → Knowledge Base or Help → Known Issues →
product.
• New software features: choose Help → New Features.
• Python scripting: choose Help → Python Module Overview.
• Utility programs: choose Help → About Utilities.
• Keyboard shortcuts: choose Help → Keyboard Shortcuts.
• Installation and licensing: see the Installation Guide.
• Running and managing jobs: see the Job Control Guide.
• Using Maestro: see the Maestro User Manual.
• Maestro commands: see the Maestro Command Reference Manual.
Contacting Technical Support
If you have questions that are not answered from any of the above sources, contact Schrödinger
using the information below.
Web:
E-mail:
Mail:
Phone:
Fax:
FTP:
http://www.schrodinger.com/supportcenter
help@schrodinger.com
Schrödinger, 101 SW Main Street, Suite 1300, Portland, OR 97204
+1 888 891-4701 (USA, 8am – 8pm Eastern Time)
+49 621 438-55173 (Europe, 9am – 5pm Central European Time)
+1 503 299-4532 (USA, Portland office)
ftp://ftp.schrodinger.com
Generally, using the web form is best because you can add machine output and upload files, if
necessary. You will need to include the following information:
•
•
•
•
122
All relevant user input and machine output
Glide purchaser (company, research institution, or individual)
Primary Glide user
Installation, licensing, and machine information as described below.
Schrödinger Software Release 2015-2
Getting Help
Gathering Information for Technical Support
The instructions below describe how to gather the required machine, licensing, and installation
information, and any other job-related or failure-related information, to send to technical
support. Where the instructions depend on the profile used for Maestro, the profile is indicated.
For general enquiries or problems:
1. Open the Diagnostics panel.
•
•
•
•
Maestro: Help → Diagnostics
Windows: Start → All Programs → Schrodinger-2015-2 → Diagnostics
Mac: Applications → Schrodinger2015-2 → Diagnostics
Command line: $SCHRODINGER/diagnostics
2. When the diagnostics have run, click Technical Support.
A dialog box opens, with instructions. You can highlight and copy the name of the file.
3. Upload the file specified in the dialog box to the support web form.
If you have already submitted a support request, use the upload link in the email response
from Schrödinger to upload the file. If you need to submit a new request, you can upload
the file when you fill in the form.
If your job failed:
1. Open the Monitor panel, using the instructions for your profile as given below:
• Maestro/Jaguar/Elements: Tasks → Monitor Jobs
• BioLuminate/MaterialsScience: Tasks → Job Monitor
2. Select the failed job in the table, and click Postmortem.
The Postmortem panel opens.
3. If your data is not sensitive and you can send it, select Include structures and deselect
Automatically obfuscate path names.
4. Click Create.
An archive file is created, and an information dialog box with the name and location of
the file opens. You can highlight and copy the name of the file.
5. Upload the file specified in the dialog box to the support web form.
If you have already submitted a support request, use the upload link in the email response
from Schrödinger to upload the file. If you need to submit a new request, you can upload
the file when you fill in the form.
Glide 6.7 User Manual
123
Getting Help
6. Copy and paste any log messages from the window used to start the interface or the job
into the web form (or an e-mail message), or attach them as a file.
• Windows: Right-click in the window and choose Select All, then press ENTER to
copy the text.
• Mac: Start the Console application (Applications → Utilities), filter on the application that you used to start the job (Maestro, BioLuminate, Elements), copy the text.
If Maestro failed:
1. Open the Diagnostics panel.
• Windows: Start → All Programs → Schrodinger-2015-2 → Diagnostics
• Mac: Applications → SchrodingerSuite2015-2 → Diagnostics
• Linux/command line: $SCHRODINGER/diagnostics
2. When the diagnostics have run, click Technical Support.
A dialog box opens, with instructions. You can highlight and copy the name of the file.
3. Upload the file specified in the dialog box to the support web form.
If you have already submitted a support request, use the upload link in the email response
from Schrödinger to upload the file. If you need to submit a new request, you can upload
the file when you fill in the form.
4. Upload the error files to the support web form.
The files should be in the following location:
• Windows: %LOCALAPPDATA%\Schrodinger\appcrash
(Choose Start → Run and paste this location into the Open text box.)
Attach maestro_error_pid.txt and maestro.exe_pid_timestamp.dmp.
• Mac: $HOME/Library/Logs/CrashReporter
(Go → Home → Library → Logs → CrashReporter)
Attach maestro_error_pid.txt and maestro_timestamp_machinename.crash.
• Linux: $HOME/.schrodinger/appcrash
Attach maestro_error_pid.txt and crash_report_timestamp_pid.txt.
If a Maestro panel failed to open:
1. Copy the text in the dialog box that opens.
2. Paste the text into the support web form.
124
Schrödinger Software Release 2015-2
Index
A
acceptor constraints........................................... 62
Active Site Residues dialog box ....................... 33
adding hydrogens .............................................. 24
amide bonds, rotation around............................ 52
annealing ........................................................... 10
applyhtreat utility ........................................ 24
B
batch processing of ligands............................. 115
binding affinity.................................................. 10
bonds
amide, rotation............................................ 52
covalent, to receptor ................................... 21
rotatable ........................................................ 8
C
ChemScore........................................................ 10
close contacts .................................................... 19
conformational search......................................... 7
conformations
core ............................................................... 8
receptor ................................................. 15, 28
ring.............................................................. 52
constraints ......................................................... 12
acceptor....................................................... 62
application of .............................................. 60
changing name of ................................. 37, 39
definitions ................................................... 34
donor........................................................... 62
feature definitions ............................... 64, 105
feature file................................................. 106
feature type ................................................. 64
file............................................................... 35
keywords................................................... 105
ligand core .................................................. 58
markers ....................................................... 61
metal atom .................................................. 37
metal coordination ...................................... 39
multiple....................................................... 67
NOE............................................................ 36
number of atoms matching ......................... 66
optional ....................................................... 62
point at which tested................................... 67
positional .................................................... 36
satisfaction of ............................................. 77
torsional ...................................................... 68
Constraints tab, Ligand Docking panel ............ 61
Constraints tab, Receptor Grid Generation
panel .............................................................. 34
H-bond/Metal subtab.................................. 38
Metal Coordination subtab ......................... 40
Positional subtab ........................................ 35
conventions, document ..................................... vii
core conformations ............................................. 8
Core tab, Ligand Docking panel ....................... 59
core, ligand ................................................... 8, 58
Coulomb energy................................................ 11
D
diameter, ligand................................................... 9
dielectric constant, setting ................................ 57
directory
installation .................................................... 2
Maestro working..................................... 2, 95
output.......................................................... 70
utilities ........................................................ 94
distributed processing ............................... 46, 115
docking
extra-precision ............................................ 13
flexible........................................................ 51
rigid ........................................................ 8, 52
subjobs...................................................... 115
docking mode.............................................. 13, 51
docking precision.............................................. 51
docking score .................................................... 76
donor constraints............................................... 62
E
Edit Conditional Coloring dialog box............... 86
Edit Feature dialog box..................................... 64
efficiency metrics.............................................. 76
electrostatic grids .............................................. 10
Emodel .......................................................... 7, 10
energy
Coulomb ..................................................... 11
internal strain........................................ 10, 73
van der Waals ............................................. 11
energy grid score............................................... 10
ensemble pose-viewer files ............... 79, 109, 110
Glide 6.7 User Manual
125
Index
environment variable
SCHRODINGER .............................................. 1
SCHRODINGER_MAX_RETRIES ................ 109
excluded volumes
applying ...................................................... 54
defining....................................................... 42
importing from Phase ................................. 44
keywords........................................... 100, 102
penalty for occupying ................................. 54
Excluded Volumes tab, Receptor Grid Generation
panel............................................................... 43
Export Rotatable Groups dialog box ................ 84
extra-precision (XP) mode.......................... 13, 51
F
feature definitions ............................................. 64
adding SMARTS patterns to....................... 65
excluding groups from........................ 66, 105
feature type, selecting for constraint................. 64
files
ligand pose.................................................. 71
name conventions ....................................... 95
output format .............................................. 71
pose output ................................................. 70
pose viewer ................................................. 71
flexible docking .......................................... 10, 51
flexible ligands.................................................... 8
flexible receptor .................................... 14, 41, 49
force field .............................................. 10, 25, 50
format conversion, to Maestro .......................... 24
functional groups, excluding from feature
definitions .............................................. 66, 105
G
Glide XP Visualizer panel................................. 85
glide_rescore utility.................................. 110
glide_sort utility ........................................ 109
GlideScore ........................................................ 10
greedy scoring................................................... 10
grid box ............................................................. 37
grids, receptor ................................................... 10
H
high-throughput virtual screening (HTVS)
docking mode ................................................ 51
126
Schrödinger Software Release 2015-2
hydrogen bonds
constraint types...........................................
defining constraints ....................................
geometric criteria for ..................................
intramolecular.............................................
receptor atom sites......................................
resolving clashes ........................................
hydrogen treatment ...........................................
hydroxyl groups, rotation of .............................
62
38
37
54
37
19
24
41
I
identical chains ................................................. 17
impact command........................................... 115
Induced Fit protocol.......................................... 14
interactions, viewing per-residue ...................... 83
internal strain energy ........................................ 10
penalty for .................................................. 73
reducing with postminimization................. 72
J
job control commands....................................... 95
job incorporation............................................... 46
Job Settings dialog box ..................................... 46
grid generation jobs .................................... 26
jobs, restarting................................................. 109
L
license, LigPrep ................................................ 22
Ligand Docking panel
Constraints tab............................................ 61
Core tab ...................................................... 59
Ligands tab ................................................. 48
Output tab ................................................... 71
Settings tab ................................................. 50
Torsional Constraints tab............................ 69
ligand pose file.................................................. 71
ligands
constraint features ...................................... 63
diameter ........................................................ 9
efficiency metrics ....................................... 76
file format ................................................... 47
flexible.......................................................... 8
pose, definition ............................................. 7
rigid .............................................................. 8
skipped, file of .......................................... 108
Ligands tab, Ligand Docking panel.................. 48
Index
LigPrep........................................................ 22–23
limits
number of atoms in ligand.......................... 49
number of poses per ligand......................... 72
number of rotatable bonds .......................... 49
total number of poses ................................. 72
M
metal-ligand interaction
constraints................................................... 37
state penalties ............................................. 53
metalloproteins, charge and protonation states. 21
metals
coordination of ligands to........................... 21
covalent bonds to protein............................ 18
use of Epik state penalties .......................... 53
minimization
of docked poses .......................................... 72
of rough-scored poses................................. 57
multimeric protein structures ............................ 17
multiple processors, running on...................... 115
N
New Excluded Volume dialog box.................... 43
New Pattern dialog box..................................... 65
New Position/NOE dialog box.......................... 36
nuclear Overhauser effect ................................. 34
strain ........................................................... 75
torsions ............................................... 52, 112
per-residue interactions
calculating .................................................. 73
keywords................................................... 108
viewing ....................................................... 83
Per-Residue Interactions panel ......................... 83
pose viewer file ................................................. 71
Pose Viewer panel............................................. 80
poses ................................................................... 7
bypassing initial elimination ...................... 57
comparing between runs............................. 82
number written out ..................................... 72
output file ........................................... 70, 107
post-docking minimization......................... 73
refining input .............................................. 52
RMSD from input geometry ...................... 72
scoring input ............................................... 53
viewing ....................................................... 79
viewing by ligand ....................................... 81
product installation ......................................... 122
properties, output .............................................. 76
protein
adjustment of structure ............................... 18
misprotonation of ....................................... 19
protein preparation, overview ........................... 17
protonation state................................................ 21
R
O
Output - Advanced Settings dialog box ............ 74
output file format .............................................. 71
Output tab, Ligand Docking panel.................... 71
overview of protein preparation ........................ 17
P
para_glide utility ........................................ 115
partial charges
ligand .......................................................... 48
receptor ....................................................... 29
threshold for nonpolar atoms...................... 28
penalties
buried polar groups..................................... 11
excluded volume......................................... 54
Maestro properties ...................................... 76
state (Epik) ................................................. 53
receptor
constraint sites ............................................ 12
exporting with alternative hydroxyl
orientation ............................................... 83
flexibility .................................. 14, 28, 41, 49
grids ............................................................ 10
Receptor - Advanced Settings dialog box......... 29
Receptor Grid Generation panel
Excluded Volumes tab ................................ 43
Receptor tab................................................ 27
Rotatable Groups tab .................................. 42
Site tab........................................................ 32
Receptor tab, Receptor Grid Generation panel. 27
reference ligand ................................................ 58
refining input poses........................................... 52
residues
active site .................................................... 32
for per-residue interactions......................... 73
Glide 6.7 User Manual
127
Index
missing........................................................ 18
resolving H-bond conflicts ......................... 20
with rotatable hydroxyls ............................. 41
rewards
hydrophobic................................................ 11
Maestro properties ...................................... 76
polar interactions ........................................ 11
rigid docking ................................................. 8, 52
ring conformations...................................... 52, 55
rings, 5– and 6–membered.................................. 8
RMSD
from input geometry ............................. 72, 77
from reference ligand ................................. 58
restriction on ligand core ............................ 58
rotamer groups, ligand ........................................ 8
rotatable bonds .................................................... 8
constraining ................................................ 68
rotatable groups........................................... 28, 41
Rotatable Groups tab, Receptor Grid Generation
panel............................................................... 42
S
sampling method
SP mode........................................................ 8
XP mode ..................................................... 13
Schrödinger contact information .................... 122
score
docking ....................................................... 76
energy grid.................................................. 10
GlideScore .................................................. 10
model energy .............................................. 10
per-residue interaction .......................... 73, 83
score-in-place.................................................... 53
scoring function ................................................ 10
scoring input poses............................................ 53
Settings - Advanced Settings dialog box .......... 55
Settings tab, Ligand Docking panel .................. 50
similarities, XP descriptor................................. 88
Site - Advanced Settings dialog box................. 34
Site tab, Receptor Grid Generation panel ......... 32
SiteMap binding site
selecting for grid generation....................... 28
specifying box size and center for .............. 31
SMARTS patterns
for feature definitions ................................. 65
for torsional constraints .............................. 68
128
Schrödinger Software Release 2015-2
SP mode ............................................................ 51
standard-precision (SP) docking mode ....... 13, 51
strain correction ................................................ 73
parameters for............................................. 75
strain energy...................................................... 10
structures
format conversion ....................................... 24
requirements for docking ........................... 21
subjobs
docking ....................................................... 46
pose files............................................. 75, 108
restarting failed........................................... 46
subjobs, Glide docking ................................... 115
subset test.......................................................... 10
symmetry-equivalent atoms .............................. 37
T
Torsional Constraints tab, Ligand Docking
panel .............................................................. 69
torsions
biasing sampling of .............................. 51, 52
constraining ................................................ 68
constraints keywords ................................ 106
defining restrictions on ............................. 111
penalties.................................................... 112
U
utilities
applyhtreat ............................................ 24
glide_rescore...................................... 110
glide_sort ............................................ 109
location of................................................... 94
para_glide ............................................ 115
V
van der Waals grids........................................... 10
van der Waals radii, scaling of.................... 14, 28
X
XP descriptors, writing ..................................... 53
XP mode ..................................................... 13, 51
XP Visualizer .................................................... 84
table color schemes .................................... 86
table navigation .......................................... 87
120 West 45th Street
17th Floor
New York, NY 10036
101 SW Main Street
Suite 1300
Portland, OR 97204
245 First Street
Riverview II, 18th Floor
Cambridge, MA 02142
8910 University Center Lane
Suite 270
San Diego, CA 92122
155 Gibbs St
Suite 430
Rockville, MD 20850-0353
Dynamostraße 13
D-68165 Mannheim
Germany
Zeppelinstraße 73
D-81669 München
Germany
Potsdamer Platz 11
D-10785 Berlin
Germany
Quatro House
Frimley Road
Camberley GU16 7ER
United Kingdom
8F Pacific Century Place
1-11-1 Marunouchi
Chiyoda-ku, Tokyo 100-6208
Japan
No. 102, 4th Block
3rd Main Road, 3rd Stage
Sharada Colony
Basaveshwaranagar
Bangalore 560079, India
SCHRÖDINGER
®
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertising