overview of subworkflow sketches

overview of subworkflow sketches
Wings Drugome: Analysis and extraction of the subworkflows -- Daniel Garijo & Yolanda Gil
Date: 07/09/2011 (MM/DD/YYYY)
(Started:07/05/2011)
Analysis of the “Methods” section of the paper: Templates.
For each “Method”:
●
●
●
●
●
Tools that participate. Analysis
Use of databases/databanks. Analysis.
Figure: how could this be a sub-workflow? (Draft aproximation)
How this “Method” could be linked to other methods (subworkflows)?
Other questions
For each Tool:
The analysis of the tools is mainly a:
● Feature description of the tool
● The process followed by the tool to produce the expected outcome
● What are the inputs and outputs needed for the tool to work?
● What are the parameters needed for the tool to work properly?
● Can it be run through the command line?
● Is it open source?
● Do I need to install it locally or is it a package/library that I can download and run?
● Example of usage.
For each Dataset/Databank:
●
●
●
●
●
●
Main features of the dataset/databank
Is it open access ?
How do I query the dataset/ bank? (Is there a web service, or something like that?)
Can it be accessed remotely?
Example of query.
Usage in the step
*************************************************************************************************************
*************************************************************************************************************
********
Structural coverage of the M.tb proteome
●
●
●
■
■
■
■
■
■
■
■
■
■
■
■
Description of the method: In this step takes place the selection of the proteins and the
homology models to be used in the study.
Tools that participate. Analysis
○ None.
Use of databases/databanks. Analysis.
○ RCSB PDB: Research Collaboratory for Structural Bioinformatics Protein Data
Bank (http://www.pdb.org/pdb/home/home.do)
Features: The Protein Data Bank (PDB) archive is the single worldwide
repository of information about the 3D structures of large biological
molecules, including proteins and nucleic acids.
Is it Open?: Yes (free of all copyright restrictions and made fully and freely
available for both non-commercial and commercial use)
How to query it?: the website
(http://www.pdb.org/pdb/search/advSearch.do), ftp server (supporting ftp
and rsync access (ftp://ftp.wwpdb.org)), Web Services and RSS feeds.
Can it be accessed remotely? YES:
http://www.pdb.org/pdb/software/rest.do
Query example: In the website, you can enter a protein ID and search the
results. It can be also visualized in 3D with some java applets.
Usage in this step: Find which proteins in the M.tb proteome have solved
structures in the RCSD PDB.
○ ModBase (homology models): http://modbase.compbio.ucsf.edu/modbasecgi/index.cgi
Features: queryable database of annotated protein structure models.
MODBASE contains theoretically calculated models, which may contain
significant errors, not experimentally determined structures.
Open? Yes. Users of ModBase must cite a specific paper in their
publications.
How to query it: ModBase Search form:
http://modbase.compbio.ucsf.edu/modbase-cgi/index.cgi . Login
requested for advanced features.
Can it be accessed remotely? (I haven’t found any web services to
access it)
Query example: (No need, since it is donde with the graphical interface)
Usage in this step: ModBase is used as the database where the
homology models for the M.tb proteome have been retrieved. Each model
is assigned a ModPipe Protein Quality Score, and if this score is greater
than 1.1 then the model can be considered reliable. MPQS is a composite
score comprising sequence identity to the template, coverage, and the
three individual scores evalue, z-Dope and GA341. We consider a MPQS
of >1.1 as reliable. A total of 1446 reliable homology models have been
selected.
○ Table S4 in the Supporting Information section:
■
●
●
Information about the solved M.tb structures used in the TB-drugome. For
each protein, the gene name (if available), gene accession number,
protein name and corresponding PDB codes are given.. Link to the sheet:
http://www.ploscompbiol.org/article/fetchSingleRepresentation.action?uri=
info:doi/10.1371/journal.pcbi.1000976.s009
○ Table S5 in the Supporting Information section: Information about the M.tb
homology models used in the TB-drugome. For each homology model, the
ModBase model code is given, as well as the gene accession number, gene
name and description of the M.tb protein. N.B. Link:
http://www.ploscompbiol.org/article/fetchSingleRepresentation.action?uri=info:doi
/10.1371/journal.pcbi.1000976.s010
Figure: how could this be a sub-workflow? (Draft aproximation)
Steps of the task:
1. Look up in PDB the proteins of the M.tb proteome with solved structures.
2. Select the multiple structures available for the proteins.
3. The outcome of this is table S4.
4. Look up in ModBase the homology models of the proteome.
5. Assign to each of them the MPQS.
6. Select the models with MPQS>1.1
7. The outcome of this process is table S5.
How this “Method” could be linked to other methods (subworkflows)? The outcomes
produced by this subworkflow are 2: table S4 and table S5.
●
Additional questions:
○ How is this step done?, manually?
○ The access to ModBase, how is it done? I just downloaded the flat files
from the database and parsed them using a script
○ The MPQS assignation is done manually or there is an additional
script/tool? The MPQS was extracted from the flat files
Identification of FDA-approved drug binding sites
●
●
●
■
■
■
■
■
■
■
■
■
■
■
■
■
Description of this step: In this step takes place the selection of the approved drugs
binding sites in the United States and Europe used in the study.
Tools that participate. Analysis.
○ None
Use of databases/databanks. Analysis.
○ Food And Drug Administration Orange Book
Main features of the dataset/databank:
http://www.accessdata.fda.gov/scripts/cder/ob/default.cfm. Drugs for
human use.
Is it open access? Yes
How do I query it? They have a Web page in
http://www.accessdata.fda.gov/scripts/cder/ob/default.cfm
Can it be accessed remotely? There is no need: it can be downoladed
from http://www.fda.gov/Drugs/InformationOnDrugs/ucm129689.htm
Example of query: N/A (direct usage of the web service)
Usage in the step: Query to retrieve the drugs for human use in the US
○ European Medicines Agency
Main features of the dataset/databank: http://www.ema.europa.eu/ema/.
Drugs for human use in Europe. Monitorized safety of the drugs in the
dataset.
Is it open access? Yes
How do I query the dataset bank? Via the frontend exposed at
http://www.ema.europa.eu/ema/
Can it be accessed remotely?There is no need. It can be downloaded
from:
http://www.ema.europa.eu/ema/index.jsp?curl=pages/document_library/d
ocument_listing/document_listing_000312.jsp&murl=menus/document_lib
rary/document_library.jsp&mid=WC0b01ac0580022517
Example of query (N/A, since it is done providing the medicine name at
the web page)
Usage in the step: Query to retrieve the available drugs for human use in
Europe.
○ PubChem
Main features of the dataset/databank: http://pubchem.ncbi.nlm.nih.gov/.
Provides information on the biological activities of small molecules, and it
■
■
■
■
■
○
■
■
■
■
■
■
○
■
■
■
■
■
■
○
is organized as three diferent linked datasets (Substance, Compbound
and BioAssay)
Is it open access? Yes
How do I query the dataset bank? Via its frontend
(http://pubchem.ncbi.nlm.nih.gov/). Lists of IDs can also be queried:
http://pubchem.ncbi.nlm.nih.gov/idexchange/idexchange.cgi
Can it be accessed remotely? ftp: ftp://ftp.ncbi.nlm.nih.gov/pubchem/
(bulk data download)
Example of query: N/A.
Usage in the step: Used for extraction of the names of the active
ingredients of the drugs.
DrugBank
Main features of the dataset/databank: Combines detailed drug (i.e.
chemical, pharmacological and pharmaceutical) data with comprehensive
drug target (i.e. sequence, structure, and pathway) information.
Is it open access? Yes
How do I query the dataset bank? Via the web page:
http://www.drugbank.ca/extractor. Results can be shown in various
formats (HTML, CSV). Text query is also possible:
http://www.drugbank.ca/search/advanced
Example of query: N/A.
Can it be accessed remotely?It can be downloaded:
http://www.drugbank.ca/downloads
Usage in the step: Used for extraction of the names of the active
ingredients of the drugs.
ChEBI
Main features of the dataset/databank: http://www.ebi.ac.uk/chebi/.
Stands for Chemical Entities of Biological Interest (ChEBI). Centered in
small molecular entity groups. It also provides an ontological
classification, capturing the relationships between entity classes and their
parents/childs.
Is it open access? Yes
How do I query the dataset? Via the web page:
http://www.ebi.ac.uk/chebi/init.do or
http://www.ebi.ac.uk/chebi/advancedSearchForward.do (advanced
search)
Can it be accessed remotely?It can be accessed via ftp:
ftp://ftp.ebi.ac.uk/pub/databases/chebi/ or downloaded from the next link:
http://www.ebi.ac.uk/chebi/downloadsForward.do
Example of query: N/A.
Usage in the step: Used for extraction of the names of the active
ingredients of the drugs.
Table S1 in the Supporting Information Section: Information about the
approved drug binding sites used in the TB-drugome. For each drug, its name,
●
PDB ligand code, isomeric SMILES string and known targets are listed, and the
PDB codes of the protein structures with which it has been crystallized are given.
Link:
http://www.ploscompbiol.org/article/fetchSingleRepresentation.action?uri=info:doi
/10.1371/journal.pcbi.1000976.s006
Figure: how could this be a sub-workflow? (Draft aproximation)
Steps of the task:
1. Search for drugs approved by the US and Europe in FDA and EMEA.
2. Obtain the names of the active ingredients of the drugs.
3. Map the compounds in PubChem, DrugBank and ChEBI.
4. Remove nutraceuticals and prodrugs.
●
●
5. Use InChI keys to map the remaining compunds to proteins structures in PDB,
excluding non-protein crystal structures.
6. Outcome of the process is table S1.
How this “Method” could be linked to other methods (subworkflows)? The outcome of
this subworkflow is table S1. Still to be seen how it is connected to others.
Other questions:
○ Are the InChI keys already provided, or are they obtained using their
software (http://www.iupac.org/inchi/release102.html)
Comparison of ligand binding sites using SMAP
●
●
■
■
■
■
■
■
■
■
Description of the step: In this step SMAP tool is used to compare the binding sites of of
the 749 protein structures + 1446 homology models with the 962 binding sites of the 274
approved drugs. The outcome is a p-value for each pair compared.
Tools that participate. Analysis
○ SMAP
Features of the tool: designed for the comparison and the similarity
search of protein three-dimensional motifs independent on the sequence
order.
The process followed by the tool to produce the expected outcome:
Based on a sequence order independent profile-profile alignement
(SOIPPA algorithm). It also uses the MWSG algorithm to align two protein
structures using a maximum weighed sub-graph.
What are the inputs and outputs needed for the tool to work? It requires 2
structures. For each it is required:
1. PDB ID or PDB File
2. Chain ID
What are the parameters needed for the tool to work properly? All the
parameters are specified at:
http://nbcr.sdsc.edu/pub/wiki/index.php?title=SMAP_Opal_Services#Prog
rammatic_Access
Can it be run through the command line? It offers a WS, accesible
programatically:
http://nbcr.sdsc.edu/pub/wiki/index.php?title=SMAP_Opal_Services#Prog
rammatic_Access. It can also be downloaded.
Is it open source? Yes
Do I need to install it locally or is it a package/library that I can download
and run? I can download it from
http://funsite.sdsc.edu/scb/smap/document.html#%20Installation
(Installation instructions)
Example of usage: (from the WS python GenericServiceClient.py \
-l http://kryptonite.nbcr.net/opal2/services/SMAPPairComp \
-r launchJob \
-a "template_cif_chain=a template_pdb_id=1qkt
query_cif_chain=a query_pdb_id=1ohp
●
●
●
VIRTUAL_LIGAND_RANGE_CUTOFF=5.0
LIGAND_CONTACT_DISTANCE_CUTOFF=10.0
SCORE_MATRIX=McLACHLAN
TEMPLATE_LIGAND_SITE_ONLY=false
QUERY_LIGAND_SITE_ONLY=false".
Example using the local installation of SMAP:
http://funsite.sdsc.edu/scb/smap/document.html#%20Parameter%20setti
ngs
○ SOIPPA and MWSG algorithms are not described because they are used by the
SMAP tool. They are not a separate process.
Use of databases/databanks. Analysis.
○ None
Figure: how could this be a sub-workflow? (Draft aproximation)
Steps of the task:
1. Takes as input table S4 (749 proteins), the 1446 homology models from table S5
as the first parameter for SMAP. The second one is the 962 bindings for the 274
drugs (table S1). There is also a script to call the pairs of the proteins.
2. Compare the binding sites with the approved drugs in an all-against-all manner.
3. For each pairwise comparison, a P-value is produced (the p-value is the
probability of obtaining a test statistic at least as extreme as the one that was
actually observed, assuming that the null hypothesis is true). This p-value
represent the significance of the binding.
4. Where is the output located? 1 FILE per each pair. Additional step to extract
all the values and drop them into a table. They have a script, so we would
need it. PERL scripts.
How this “Method” could be linked to other methods (subworkflows)? It uses the
outcome of the previous 2 sub-workflows as the input to make the comparison and
produce the p-values. I assume that the output from the SMAP tool is a list of pvalue result, one for each comparison (this is not specified in the paper). It would
be a list of 2195*962 = 2.111.590 entries (or a table of 2195 rows and 962 columns)
Comparison of global protein structures using FATCAT
●
●
■
■
■
■
■
■
■
●
Description of the step: In this step takes place the use of FATCAT
Tools that participate. Analysis
○ FATCAT
Feature description of the tool: Flexible structure AlignemenT by Chaining
Aligned fragment pairs allowing Twists. It is used to report the overall
similarity between 2 structures, using a p-value to measure it. P-value
less than 0.05 means that they are similar.
The process followed by the tool to produce the expected outcome. The
user fills a form providing either the PDB code, file in PDB format or
SCOP domain code of the two resources to be aligned and submits it to
the server. The response is a p-value with the similarity and some
additional outcomes from the server. Example:
http://fatcat.burnham.org/fatcat/examples/1ufhA_1gheA/. Form available
at: http://fatcat.burnham.org/fatcat-cgi/cgi/fatcat.pl?-func=pairwise
What are the inputs and outputs needed for the tool to work? PDB code,
file in PDB format or SCOP domain code of the 2 entites to compare.
What are the parameters needed for the tool to work properly? No
additional parameters are required.
Can it be run through the command line?Apparently Not. But it looks like
someone just could fill the form automatically, submit it and treat the
results.
Is it open source? Yes
Do I need to install it locally or is it a package/library that I can download
and run? Not available to install it locally
Use of databases/databanks. Analysis.
○ None
●
●
●
Figure: how could this be a sub-workflow? (Draft aproximation)
Steps:
1. Query the PDB for the PDB files. Are these the same PDB files as in table S4?
2. Use FATCAT to filter the non-similar structures. Assuming it is table s4, Is each
PDB file compared to all the other ones?
3. The output would be (assuming that we are taking the 749 structures from the
PDB in table S4) a file with 794*794 = 630.436 p-values
4. Remove the pairs with high similarity (pvalue<0.05)
How this “Method” could be linked to other methods (subworkflows)?
○ It is not clear how this relates to other sub workflows. This step occurs before the
SMAP step, which uses as input the table S1? How is the list with the non similar
structures used later?
Other questions.
○ Should ask if they do this step manually or if they access some other
service to send multiple queries. Check with Sarah - I used Andreas’
JFatCat program (Java implementation of FATCAT)
○ The PDB files mentioned in the paper, is it a reference to table s4? if not,
which other PDB files are they referring to in this step?
Visualization of the protein-drug interaction network
●
●
■
■
■
■
■
■
■
■
●
■
■
■
■
■
■
●
Description of the step: This step describes the use of yEd as a Graphical editor to
visualize the protein-drug interaction.
Tools that participate. Analysis
○ yEd Graph Editor.
Feature description of the tool: Diagram editor.
The process followed by the tool to produce the expected outcome. N/A
What are the inputs and outputs needed for the tool to work? The
network from FATCAT. A Formatting is required in order to represent
the network with yEd. Cytoscape could be an alternative. Check with
Lei and Sarah if they have it. I wrote a script to generate the input for
yEd in the correct format.
What are the parameters needed for the tool to work properly?N/A
Can it be run through the command line?Yed uses a graphical interface. It
can not be run directly with the command line. In fact, it is not allowed:
Clarifies that using yEd as part of an automated process is not
allowed.
Is it open source? It is free, but not open source.
Do I need to install it locally or is it a package/library that I can download
and run? You need to install it locally
Example of usage. N/A
Use of databases/databanks. Analysis.
○ NCBI Entrez:
Main features of the dataset/databank: collection of sequences from
several sources, including translations from annotated coding regions in
GenBank, RefSeq and TPA, as well as records from SwissProt, PIR, PRF,
and PDB
Is it open access ? YES
How do I query the dataset/ bank? (Is there a web service, or something
like that?) It offers a web access to the database, filling a form or by key
words: http://www.ncbi.nlm.nih.gov/books/NBK44863/. FTP:
ftp://ftp.ncbi.nlm.nih.gov/genbank/, ftp://ftp.ncbi.nih.gov/refseq/
Can it be accessed remotely? I haven’t found any other remote
access. MANUALLY by filling a form. Maybe we can automatize this
writing a component.
Example of query. N/A
Usage in the step. To query the names of the M.tb proteins, in order to
avoid inconsistences with the naming proteins of the PDB.
Figure: how could this be a sub-workflow? (Draft aproximation)
○ No need to represent the workflow for this task (I think)
Steps:
1. (Assumption) Treat the information produced by SMAP into a table/Matrix so it
can be read directly by yEd.
●
●
2. Query the M.tb protein names to the Entres protein database.
3. Represent the graph using yEd.
How this “Method” could be linked to other methods (subworkflows)?
○ The input used for the graph representation is the same one produced by SMAP.
Other questions:
○ What is the input received by yEd? (I have assumed it is the one from SMAP)
○ Is this process necessary for the study or for just seeing the results ?
○ How is this step linked to other parts of the workflow?
Flux balance analysis
●
●
■
■
■
■
■
■
■
■
●
■
■
■
Description of the step: In this step both the in vivo essentiality and the in vitro
essentiality are calculated using different toolboxes and databases. (I assume that non
essential genes are discarded).
Tools that participate. Analysis
○ COBRA Toolbox:
Feature description of the tool.
http://opencobra.sourceforge.net/openCOBRA/Welcome.html. Constraints
Based Reconstruction and Analysis (COBRA) focuses on employing
physicochemical constraints to define the set of feasible states for a
biological network in a given condition based on current knowledge.
The process followed by the tool to produce the expected outcome. It is
used to delete single genes on the iNJ661 model.
What are the inputs and outputs needed for the tool to work? It uses the
iNJ661 model grown in 7H9 media. No additional information is
provided. How is the iNJ661 used? which functions/scripts of the
COBRA Toolbox are used?
What are the parameters needed for the tool to work properly?
Installations instructions+ usage available at :
http://www.nature.com/protocolexchange/protocols/2097#/procedure
Can it be run through the command line? Yes, because they are Matlab
scripts.
Is it open source? NO. It is designed to run with Matlab, which is not open
source. However the toolbox is free.
Do I need to install it locally or is it a package/library that I can download
and run? Local install (Matlab+Cobra toolbox download)
Example of usage. N/A, since the methods of the toolbox used for the
study are not known.
Use of databases/databanks. Analysis.
○ GSMN-TB
Main features of the dataset/databank: Web-based genome -scale
network model used to carry out the flux balance analysis.
Is it open access ?Yes
How do I query the dataset/ bank? (Is there a web service, or something
like that?) http://sysbio3.fhms.surrey.ac.uk/cgi-
bin/fba/fbapy?model=GSMN_TB-Viv&cmd=methods. A web form based
query is the only available way to use the service. It can be modified:
http://sysbio3.fhms.surrey.ac.uk/cgi-bin/fba/fbapy?model=GSMN_TBViv&cmd=fba.
Can it be accessed remotely?Aparently not. Have the authors accessed
in some other direct way? No, it was just accessed using the web.
This didn’t matter, as it was only for a small number of genes.
Example of query. http://sysbio3.fhms.surrey.ac.uk/cgibin/fba/fbapy?model=GSMN_TB-Viv&cmd=fba already shows a query. It
returns a file like this, showing the FBA:
http://sysbio3.fhms.surrey.ac.uk/cgibin/fba/fbapy?sid=sid5008843&cmd=showfba&model=GSMN_TB-Viv
Usage in the step. This model is used to calculate the essentiality
prediction under conditions optimized for in vivo growth.
■
■
■
○
■
■
■
■
■
■
●
iNJ661
Main features of the dataset/databank
Is it open access ? Yes
How do I query the dataset/ bank? (Is there a web service, or something
like that?) It can be downloaded through the next link:
http://www.biomedcentral.com/1752-0509/1/26/additional
Can it be accessed remotely? There is no need, it can be downloaded as
a .xls sheet
Example of query.N/A
Usage in the step. Used to calculate the in vitro essentiality (with the
COBRA Toolbox).
Figure: how could this be a sub-workflow? (Draft aproximation)
Steps :
1. Use GSMN-Tb to carry out the FBA computations.
2. Use the single gene knockout tool to run essentiality prediction.
3. Constrain some genes in order to simulate multiple gene knockouts.
4. Use iNJ661 as input to the COBRA Toolbox to perform single gene deletions and
determine in vitro essentiality.
●
●
How this “Method” could be linked to other methods (subworkflows)? ??????? (it is
defined as an independent step)
Other questions
Molecular docking using eHiTS
●
●
■
■
■
■
■
■
■
■
●
●
Description of the step: in this step, molecular docking to predict the binding pose and
affinity of the drug molecule to the drug proteine takes place, using eHiTS.
Tools that participate. Analysis
○ New tool for doing the docking: Autodock vina. They have scripts for
running it.
○ eHiTS Lightning
Feature description of the tool.
http://www.simbiosys.com/ehits/ehits_benefits.html (According to the web
page): Fast, accurate, full automated, customizable tool used for docking
studies.
The process followed by the tool to produce the expected outcome. Takes
the input produced by the SMAP tool. For those proteins with cofactors,
the cofactor was added as the last residue in the protein structure prior to
docking. Nothing is said about the outcome produced by the tool.
What are the inputs and outputs needed for the tool to work?
(Assumption) The significance value of the pairs analyzed with the SMAP
tool.
What are the parameters needed for the tool to work properly? ??? (I had
no access to the user manual)
Can it be run through the command line? ???? (ask the authors of the
study/ ask for a demo) Yes e.g.:
./ehits.sh -receptor receptor_file.pdb -clip clip_file -ligand ligand_file.pdb workdir . -accuracy 6 -out output_file.sdf
Is it open source?No, you have to request for a demo and then purchase
the product.
Do I need to install it locally or is it a package/library that I can download
and run? Local install.
Example of usage. N/A (there is no documentation without a registration)
Use of databases/databanks. Analysis.
○ None
Figure: how could this be a sub-workflow? (Draft aproximation)
●
●
Steps:
1. Take the output produced by SMAP.
2. For those proteins with cofactors, add the cofactor as the last residue in the
protein.
3. Parameters: search space of 10A^3, accuracy level = 6.
4. eHiTS outcome is produced as a ??? SDF file with the resulting
conformations of the molecule and their corresponding energy scores
How this “Method” could be linked to other methods (subworkflows)? Takes as input the
output produced by SMAP. Nothing is said about the outcome produced.
Other questions
○ Ask the authors about the eHiTS tool. Is it used automatically? What is the
format of the outcome? Are there any additional parameters?
○ Are there any additional processes necessary to use the tool?
Postprocessing? After running eHiTS using the command above, the
conformation with the best score (and its score) was extracted using a
script.
Network analysis
●
●
●
●
Description of the step: Consruction of a drug-target network protein graph.
Tools that participate. Analysis
○ None
Use of databases/databanks. Analysis.
○ None
Figure: how could this be a sub-workflow? (Draft aproximation)
Steps:
1.
2.
3.
4.
●
●
Get the output from SMAP step or eHiTS step.
Fit the number of targets and their connectivity to a power law distribution.
Build a graph from the drug target network.
Compute the fraction of the largest connected component by dividing the number
of proteins in the largest single linkage cluster by the total number of proteins in
the graph.
How this “Method” could be linked to other methods (subworkflows)? The input is either
obtained from the SMAP step or the eHiTS step. nothing is said about the output
produced in this step.
Other questions
○ Is this step produced manually? Ask Lei about this
○ Is the input taken from the SMAP setp, the eHiTS step or other steps?
○ What is the format of the input produced. Is it stored somewhere in a table?
Format?
○
○
Additional tools used?
What is the format of the output?
Hierarchical clustering of protein and drug binding profiles
●
●
■
■
■
■
■
■
■
Description of the step: Hierarchical cluster of the protein and drug binding profiles,
using GenePattern 2.0.
Tools that participate. Analysis
○ GenePattern 2.0:
Feature description of the tool: genomic analysis platform that provides
access to more than 150 tools for gene expression analysis, proteomics,
SNP analysis, flow cytometry, RNA-seq analysis, and common data
processing tasks.
The process followed by the tool to produce the expected outcome. I
assume it takes the result from eHiTS or SMAP as input file. What are the
inputs and outputs needed for the tool to work?
What are the parameters needed for the tool to work properly? The
parameter is the city block distance.
Can it be run through the command line? It looks that it can be directly
accessed from Java and Matlab:
http://www.broadinstitute.org/cancer/software/genepattern/tutorial/gp_pro
grammer.html?Matlab_doc#_Using_GenePattern_from_Java. Libraries
available at:
http://genepattern.broadinstitute.org/gp/pages/downloadProgrammingLiba
ries.jsf.
Is it open source? It is free, but I believe is not open.
Do I need to install it locally or is it a package/library that I can download
and run? You can either
http://www.broadinstitute.org/cancer/software/genepattern/installer/latest/i
nstall.htm (Setting up your own server), or use the website (login
required) http://genepattern.broadinstitute.org/gp/pages/index.jsf
Example of usage: JobResult result =
gpClient.runAnalysis("urn:lsid:broad.mit.edu:cancer.software.genepattern.module
.analysis:00009:5", new Parameter[]{new Parameter("input.filename", ""), new
Parameter("column.distance.measure", "2"), new
Parameter("row.distance.measure", "0"), new Parameter("clustering.method",
"m"), new Parameter("log.transform", ""), new Parameter("row.center", ""), new
Parameter("row.normalize", ""), new Parameter("column.center", ""), new
Parameter("column.normalize", ""), new Parameter("output.base.name",
"<input.filename_basename>")}); (Java)
●
●
Use of databases/databanks. Analysis.
○ None
Figure: how could this be a sub-workflow? (Draft aproximation)
○ There is no need, since it is only one step.
Steps:
●
●
1. Take the input from SMAP/eHiTS
2. Use the clustering analysis to make a hierarchy.
3. Nothing is said about the output.
How this “Method” could be linked to other methods (subworkflows)?
Other questions
Comparison of drug chemical similarity
●
●
Description of the step: 2D fingerprint similarity.
Tools that participate. Analysis
○ OpenBabel 2.1.1:
Feature description of the tool:
http://openbabel.org/wiki/Open_Babel_2.1.1 chemical toolbox designed to
speak the many languages of chemical data.
The process followed by the tool to produce the expected outcome. It
gets table S1 (drugs) (assumption) as input to calculate the similarity.
Nothing is said about the output. Maybe is it used for another
previous step?
What are the inputs and outputs needed for the tool to work? Table S1.
What are the parameters needed for the tool to work properly?
Can it be run through the command line?Yes: The obabel command line
program converts chemical objects (currently molecules or reactions)
from one file format to another. The GUI interface is an alternative to
using the command line and has the same capabilities.
Is it open source? Yes
Do I need to install it locally or is it a package/library that I can download
and run? Libraries can be used from Java:
http://openbabel.org/docs/2.3.0/UseTheLibrary/Java.html
Example of usage. The link above covers them. They can be found in the
tutorials too: http://openbabel.org/wiki/Tutorial:Fingerprints
Eg: PROMPT> babel -L fingerprints
●
Use of databases/databanks. Analysis.
○ None
Figure: how could this be a sub-workflow? (Draft aproximation)
○ There is no need, since it is just one step
Steps:
1. Input: Table S1.
2. Process: OpenBabel
3. Output: ????
How this “Method” could be linked to other methods (subworkflows)? Nothing is said in
the step
Other questions
○ How is the output of this method used for other methods?
■
■
■
■
■
■
■
■
■
●
●
●
Additional steps/files:
http://funsite.sdsc.edu/drugome/TB/#Summary
Overview: connection between all the methods
Overview: connection between all the methods (FIX)
FIRST STEPS:
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

advertisement