User Manual xHLA

User Manual xHLA
User Manual
xHLA
Beta version
2010 (June)
Summary
Summary ............................................................................................................ 2
Introdution .......................................................................................................... 3
Demo limitations ................................................................................................. 3
Installing xHLA ................................................................................................... 4
Admin user and password (default) .................................................................... 4
Downloading NMDP allele code list.................................................................... 4
Downloading IMGT/HLA data ............................................................................. 5
xHLA Main Screen ............................................................................................. 5
Options tree ........................................................................................................ 7
General Workflow – Processing PRA Results .................................................... 8
General Workflow – Generating Histocompatibility Maps ................................. 10
Adding, deleting and navigating Local Database.............................................. 13
Compatibility/Histocompatibility Map ................................................................ 14
Introduction....................................................................................................... 14
The Histocompatibility map .............................................................................. 15
Alignment map ................................................................................................. 16
Eplets map ....................................................................................................... 18
The “blue eplets” algorithm – Acceptable and unacceptable mismatches. ....... 20
Eplet report ....................................................................................................... 21
Histocompatiblity maps and searches in databases ......................................... 22
Exporting the eplets map .................................................................................. 22
xHLA and the LabSystem Gen Architeture ....................................................... 22
Report BUGS ................................................................................................... 25
Manual revision ................................................................................................ 25
Introdution
xHLA is a graphical tool based on LabSystem Gen architecture which was
developed to assist researchers in pre and post transplant studies in
Histocompatibility laboratories. It provides a fast way to researchers manage
data of donors and recipients and to study the compatibility between them. The
tool has the following features:
1. Graphical and ease to use interface;
2. Data of donors and recipients are stored in a centralized and secure
relational database (Access);
3. Compatibility studies use both local and public domain data provided by
Web resources.
4. The tool generates a compatibility map in order to assist clinicians in their
transplant decisions.
5. Donors and recipient’s data stored in other systems or in spreadsheets
existing in the laboratory can be used in the compatibility studies easily.
6. Since uses LabSystem Gen architecture, new web resources and
algorithms can be added in order to improve tool functionalities.
7. Automation and integration of the HLA Matchmaker and HLA Fusion
programs.
We are improving xHLA tool in order to generate more valuable reports to works
HLA matchmaker algorithm. These changes are being done every day and new
versions will be deployed as soon as possible.
Demo limitations
The demo version has the following limitations:
1. Advanced searchers to local database are not available. These
searchers allow researchers to find the best scored potential donors to a
specific recipient or, the best recipients to a specific donor. These
searchers are performed both in local database and external databases.
2. Data stored in other databases or spreadsheets (external data) can not
be integrated to be used in data analyses.
Installing xHLA
The installation procedure is straightforward:
----------------------------------------------------------How to Install
----------------------------------------------------------1. Download file Application_xHLA.zip
2. File must be placed in directory C. Otherwise, a configuration procedure is
needed.
3. Unzip file
4. Locate file C:\Application_xHLA\xHLAb.exe
5. Double click xHLA.exe file.
Admin user and password (default)
Default user: Admin
Default password: Admin
Downloading NMDP allele code list
The NMDP allele list code must be downloaded every day. In order to download
the allele code list, please do the following procedure:
Go to menu File and choose the option Download NMDP Data. A confirmation
dialog box is shown (Figure 1). Choose Yes to confirm operation and the updated
data will be downloaded and unzipped. Please, restart application after this
operation. Depending on your web link speed, this operation can take several
minutes.
Figure 1 – Operation confirmation before downloading NMDP Data.
Downloading IMGT/HLA data
The IMGT/HLA database can be downloaded every time a new version of the
database is deployed. This operation can take several minutes, depending on
your web link speed. Please, restart application after this operation.
Go to menu File and choose the option Download IMGT/HLA Data. A
confirmation dialog box is shown (Figure 1). Choose Yes to download only
missing files. If you choose No, the entire database is downloaded and updated
locallly.
Figure 2 – Choose NO to download entire database. Choose
Yes to download missing files.
xHLA Main Screen
Figure 3 - xHLA main screen.
The screen is divided in four main areas: i. Tool box; ii. Options Tree; iii. Data
area and; iv. Main menu area. The Tool box is composed of three buttons:
Close application; Login/Logoff; and Turn on/ Turn off web link (Figure 4). The
green/red ball indicates that a link to web is not available or the link is slow at
the moment. To activate the web link, you just click once in this icon and it will
become green. The Login/logoff button opens the login dialog box to user lock
the station or to log as another user.
Figure 4 – Tool Box area.
The Main Menu Area has the same options showed in Options Tree and three
additional ones: 1. Exit, Download IMGT/HLA Data and Download NMDP Data
(File Menu). These databases are used in generation of the compatibility map.
You must download NMDP allele code list every day to have a more accurate
analyses.
The Options Tree allows user to access the main functions of the application
easily. Click over each node existing in the options tree and the Data area (right
side) shows the appropriate information.
Options tree
The options tree is a dynamic tree that provides direct access to the System
functionalities. There are three main nodes in the tree: Histocompatibility
map,
PRA
processing
and
Password
(Figure
3).
Selecting
the
Histocompatibility map node, the Data Area shows the form Local Database
data (donors and recipients) repository and the compatibilities maps generated
(detailed later). The PRA processing node shows, in Data Area, the form
Process files. Using this form, PRA results generated by laboratory’s
equipments are loaded and processed. The Password node shows the form
Password that provides functionalities to change user´s password and to
create new users. Figure 5 shows the PRA processing form and form Local
Database is show in Figure 3.
Figure 5 – PRA processing and Password forms.
General Workflow – Processing PRA Results
Read this section carefully before using the xHLA program.
1. First time using xHLA, select the Input, Output and HLA
Matchmaker directories (PRA processing node):
a. Input Directory – CSV files, containing PRA results, generated by
laboratories equipments (Luminex), must be stored in this
directory, whose default value is c:\Application_xHLA\Input. If you
wish, click “Change…” button to set a new directory path (Figure
5).The PRA files stored in this directory are shown in Available
CSV files area (Figure 5, on left). WARNING: Do not allow or copy
any other file format to the Input Directory, because xHLA can fall
into an endless loop (buggy).
b. Output Directory – It is used by xHLA as a place to temporary
storage while processing data. The default value is
c:\Application_xHLA\Output (Figure 5, bottom). Click button
“Change…” to set a new directory, if wished.
c. HLA Matchmaker Directory – It is where HLA Matchmaker
programs must be stored (Figure 5, bottom). Before copy a version
of the HLA Matchmaker program to this directory, be sure that the
PANEL sheet was properly filled up and that information about
beads/lot were correctly inputted (xHLA only accepts Omni
Lambda). Once a copy of the HLA Matchmaker program is
coppied to this directory, they can be added to xHLA System:
i. Addind a new HLA Matchmaker program: Click “Add...”
button to load a new HLA Matchmaker program and follow the
sequence showed in Figure 6. During the addind process, xHLA
retrieves information existing in HLA Matchmaker spreadsheet
to determinate the lot and class (I or II) of the program.
However, it is better to set them manually or correct this
information if retrieved erroneously, after step 3 of Figure 6. A
new program has to be loaded every time version or lot
information (PANEL sheet) change.
Figure 6 - Adding an HLA Matchmaker program. Follow steps 1
to 4. After clicking button OK ([4]), several minutes can be
needed. Please, wait.
ii. Deleting a program (Figure 6, top): Click “Del..” button to
delete the selected program.
2. Select the appropriate program:
3. Select the desired CSV file generate by LUMINEX
equipment with PRA results.
Figure 7 – Processing PRA results.
If the CSV file is a valid one, the data is showed in the matrix showed in
Figure 7. Click the Process button (on detail) to processe file. Once a
CSV file is processed, the PRA results loaded from CSV files are
visualized in the form Local Database (Histocompatibility map node). A
record in the local database is added for each recipient processed for the
first time.
WARNING: xHLA uses internally a unique number named Laboratory
Unique Number or LUN (Figure 3, Data Area). This number is used to
identify a donor/recipient internally by the System and it is a sequential
unique number that is created for each patient (social security number
can be used as LUN, for example). The files containing PRA results must
be produced considering this important pre-requisite. The sample
description must contain the patient name and this number (one number
per patient). The description must set as shown below:
Patient name[specialCaracter]laboratoryUniqueNumber
Where [specialCaracter] can be: #, _, @. Avoid using ; or ,. Usage
example: JohnSmith_9431
In LIB laboratory, it is used the underscore (_) as the special
character
If another character is going to be used by Lab, please inform the
System in the text edit box named “Char between name and Lab.
Unique #” (Figure 7, top).
4. Search the recipient in the Local Database form to check
PRA results, after processing operation. To access this
form, click the Histocompatibility map node (Options
Tree) and locate a recipient/donor, as explained in “General
Workflow – Generating Histocompatibility Maps” section.
General Workflow – Generating Histocompatibility Maps
A Histocompatibility map can be generated if the donor and recipient’s HLA
typings are available. However, maps generated to highly sensitized patients
require PRA results. To load PRA results related to a patient, see “General
Workflow – Processing PRA Results” section. The steps to generate a
histocompatibility map are described below:
1. Search the recipient.
a. Click the Histocompatibility map node (Options Tree).
b. The form Local Database (Figure 11) is shown. There are two
main areas: “Donors and recipient’s area” (top) and
“Histocompatibilty study’s area” (bottom).
c. Click a field of the form Donors and recipient’s area. This field
will be considered by the System as parameter in searchers. For
instance, click the Description/name field if you desire to search
for a record using its description information as search parameter.
d. Click the Locate text box (on top) and it will become Search for
text box.
e. Type information in the Search for text box that will be used as
search parameter. (Example: type JOHN, if Description was
selected as parameter).
f. Click the button Locate (In the example, all records containing the
word JOHN are returned by the System).
g. If one or more records match, they will be presented to User. If
more than one is retrieved, click the list of donors and recipients
tab to see all at once. (In the case of the example, JONH
something or something JOHN will be retrieved.
2. Type the HLA typing information in the HLA typing area
(Figure 11). Warning: Some allele codes are not recognized
by HLA Matchmaker program and, in these cases, the
most frequent in population are used instead. During
allele codes typing, the System presents them in green
color, if the typed allele is valid. Invalid alleles are shown
in black color (Figure 9)
Figure 8 – Typing na allele code. The list of codes is retrieved from
IMGT/HLA database.
3. Define the Cutoff. If PRA results are available to the
selected patient, they are presented in the PRA results tab
(Figure 9) in a HLA-Fusion like report.
Figure 9 – PRA results of a patient reported to User.
4. If a potential donor is available but not include yet, add
him/her into Local Database (see Adding, deleting and
navigating and Local Database section). If he/she is already in
database, goto step 5. If no donor is available, go to step
6.
5. If a potential exists and he/she was stored previously in
database, select the “Potential donors tab” in the
“Histocompatibilty study’s area” (Figure 10). Click add button
[1], and locate the donor in the dialog window “Search”.
Type the name of the desired donor in the “Search for”
edit box and press “Enter”. The donors retrieved are listed
in the matrix [3]. Choose one and click “OK” button [4] to
add him/her. Click “Cancel” button if you decided to abort
operation.
Figure 10 – Steps to add a potential donor to a recipient.
6. Click the “Histocompatibility map” tab to generate the
histocompatibility map.
Adding, deleting and navigating Local Database
Patient and donor’s data are stored in a Local Database (Local repository).
Thos section explains how to maintain data stored in it.
Adding:
1. Click the Histocompatibility map node (Option trees). The local database
form will be available as shown in Figure 11. Click the icon
in the tool
bar. This will add a new (blank) record.
Figure 11 – Local database form.
2. Type the information of the donor/recipient. The HLA typing is stored in
fields A, B, Cw, DRB1 and son on. Haplotype 1 is the upper line and, the
haplotype 2 is the lower line. The Description/name field is required.
3. Click the icon
Deleting:
to save the information.
Locate the record (see step 1 in General Workflow – Generating
Histocompatibility Maps section). Click the delete button
.
Navigating:
To navigate, use buttons prior and next
.
Canceling:
To cancel an input: click cancel button
.
Compatibility/Histocompatibility Map
Introduction
xHLA generates a histocompatibility map to assist researchers and clinicians in
their pre- and post-transplantations decisions. The map is divided in two major
parts and, each part, can be visualized in different level of details (). The data to
generate a histocompatibility map is retrieved from several sources, including:
a. Laboratory private data: Luminex, donor and recipient’s clinical and
genetic data and spreadsheets.
b. Public database/web site data: IMGT/HLA database and NMDP site.
c. HLA Matchmaker program.
One compatibility map is generated for each donor-recipient pair and the
recipient can be a highly sensitized one. To illustrate its use, the HLA typing in
Figure 12
will be considered to generate a Class I map. The PRA results and any
other data needed to simulate this example are available in www.ufpi.br/LIB
web site (xHLA programa download).
Recipient
Donor
A*3301
A*2402
A*6801
A*3301
B*1516
B*3501
B*7801
B*7801
Figure 12 – HLAtypings used to generate the Histocompatibility map of Erro! Fonte de
referência não encontrada.
The Histocompatibility map
The input data to generate a compatibility map are: The recipient and donor’s
HLA typings, the PRA results and cutoff. Changing any of these variables, the
resulting map also will be different. The map showed in Figure 13 is a complete
one because the four inputted variables were provided. If only HLA typings are
provided, the Eplets map is not generated, because PRA results are needed to
generate it. On the other side, if the donor’s HLA typing is not provided, the
Alignment map is not generated, because both HLA typings are needed to
generate it. The histocompatibility map is the eplet and alignment maps
together and, information existing in one, complements the data existing in the
other.
Figure 13 – A complete histocompatibility map with two parts: Eplets map [1] and
Alignment map [2].
Eplets maps are generated based on HLA Matchmaker program and eplets
data and they are important to highly sensitized patients. The Alignment map is
generated using data provided by IMGT/HLA database and, its advantage, lays
on the fact that PRA results are not needed to generate it.
Alignment map
The alignment map is generated, using the donor and recipient’s HLA typings,
as follows:
a. For each locus, the allele codes of the donor-recipient pair are compared
to each other and, if they don’t match, their nucleotide sequences are
retrieved from IMGT/HLA database and aligned.
b. The alignment mismatches are counted and summarized in the
Alignment map (Figure 13, [2]).
c. Each line of the Alignment map is summary of the alignment performed.
The details of this alignment are showed in detail by the Detailed
alignment map(Figure 14). This detailed map contains three matrixes of
data: Alignment matrix[1], Mismatches matrix [2] and Number of
mismatches by exon matrix [3].
Figure 14 – Details of the alignment between A*6801 and A*3301.
The Alignment matrix shows the alignment of the nucleotide sequences
retrieved from IMGT/HLA. This matrix has four or more lines: Nucleotide
position, recipient’s allele code, one or more donor’s allele codes and the Exon
number. If one position is pink colored, it means that there is a mismatch in that
position. If an asterisk is showed in one position, it means that the nucleotide is
unknown in that place and they both are counted as mismatches. The
Mismatches matrix shows the mismatching nucleotides found during alignment
process and, the Number of mismatches by exon (NME) matrix summarizes the
Mismatches matrix, exon by exon. The Aligment Map showed in Figure 13 is the
union of all Number of mismatches by exon matrixes generated to the donorrecipient pair.
The generation of the Alignment map using NMDP codes as input is also
possible, however the xHLA System needs to retrieve allele list codes from the
NMDP web site and combine them. The result is an Alignment map showing all
alignments possible. Figure 15 is an example. The donor’s A*01CD creates two
lines in the Alignment matrix because the allele list codes returned to this
NMDP code are A*0103 and A*0104. So, the A*2633 (recipient) is aligned to
each allele of the list.
Figure 15 – Alignment details usign A*2633 and A*01CD (NMDP code). The allele code list
returned to A*0CD is (A*0103 and A*0104N).
The alignment in the IMGT/HLA website can also be visualized on line easily,
just clicking the button
. This operation executes an integrated browser as
show in Figure 16. Notice that xHLA uses old nomenclature to generate maps,
but since nomenclature has changed in april, during web sites queries, they are
translated to new nomenclature. The xHLA system still uses the old
nomenclature because HLA MAtchmaker program was not updated until now.
Figure 16 – IMGT/HLA website integrated to xHLA graphical interface.
The best Alignment map shows the columns exon 2 and exon 3 with zero
mismatches because clinically has been reported that mismatches in these two
exons increase the chances of graft failures. During searches in waitlists or
databases, the xHLA System tries to find a donor-recipient pair with no
mismatches in these two exons.
Eplets map
To generate an Eplet map, the minimum inputted data are: Recipient’s HLA
typing, PRA results (Luminex csv files) and the eplets for each HLA molecule. In
molecular terms, a particular HLA molecule may present non-linear sequences
of amino acids (eplets) potentially shared by other HLA molecules. HLA
molecules that share the same eplets, for which the patient has preformed
antibodies and can be potentially recognized, contraindicate transplantation. On
the other hand, eplets that are not recognized by preformed antibodies in the
renal transplant recipient, in theory, offer no danger to the transplant. As a
result, any HLA molecule consisting of only eplets not recognized by the
antibodies of the patient are acceptable for transplantation. Such molecules are
now known as HLA acceptable mismatches. The elucidation of involved eplets
can be achieved by running the HLA Matchmaker (HMM) algorithm and, the
xHLA generates the Eplets map based on outputs generated by this program.
The eplets map generation in based on Cutoff value. These values will
determinate which HLA molecules are positively recognized and negatively
recognized by antibodies. So, the cutoff generates two main sets of data in the
PRA results: POS and NEG. The determination of the cutoff is crucial in results
and, the xHLA produces a report, basead on PRA results, to assist the
researcher in this determination (Figure 11, “Histocompatibility Studys are”). The
report shows the PRA results ordered from highest value to lowest value. The
idea is to try to determinate in which moment the Normal value decreases
rapidly (mostly cases). In the example, the cutoff was determinate as 1500
(Figure 17). Sometimes, the cutoff cannot be easily determinate and the default
initial value becomes 500.
Figure 17 –The chosen cutoff value was 1500 because the higher Dif value encountered
was 854 (beadID 17 and 63).
All values above Cutoff is added to the POS set and, all values below cutoff,
including itself is added to the NEG set. An additional set, name GRAY ZONE is
also created. This set is the PRA results whose normal values are too close to
the chosen cutoff value (up or down). xHLA uses these two sets to create the
following five Eplets sets:
Table 1 – Eplets sets provided by Eplet Map.
1. Acceptable mismatches
No pre-formed antibody against these HLA
molecules’
2. Unacceptable mismatches
There are pre-formed antibodies against
these HLA molecules
3. Gray Zone mismatches
These molecules are too close to cutoff
4. No mismatches
These molecules have the same eplets if
compared
to
recipient’s
HLA
typing
molecules.
5. Anti-donor mismatches
If a donor was provided, the differences,
between donor and recipient are presented
here.
6. All eplets
Union of the POS and NEG sets.
The “blue eplets” algorithm – Acceptable and unacceptable
mismatches.
If an eplet is in the NEG set, it means that pre-antibodies against it were not
found. The “blue eplets algorithm” paints in blue color all eplets found both in
NEG and POS sets. If all eplets of a HLA molecule in POS set becomes blue, it
means that no antibody against this molecule has been found and it is an
acceptable mismatch (Figure 18, [1]). After executing this algorithm, new two sets
are created: The acceptable and unacceptable mismatches sets. The first one
contains all molecules for whose there are no pre-formed antibody and, the
second one contains all HLA molecules for whose there are pre-formed
antibodies against then. Shortly, the acceptable mismatches are the NEG set +
molecules of the POS set whose all eplets are in NEG set too. The
unacceptable mismatches are the POS set - blue molecules of the POS set.
After executing this algorithm, the cutoff can be or not be upgraded. See in
Figure 18
that if, the eplets 44KM and 152RW were in NEG set also, the cutoff
would be upgraded to 2762. However, because they are not in NEG set, the
cutoff remains 1500 (44KM and 152RW are positives).
Figure 18 – The “dance of eplets” algorithm. 65QKR was found both in NEG and POS
sets.
Another important analysis is to see if the eplet in both POS set and NEG set
appears only in gray zone. The gray zone is determinate by user and is a
values around cutoff value. By default, xHLA determinates the gray zone as
Cutoff value +/- 100. In the example, gray zone ranges from 1400 to 1600
(normal value). Eplets found as neg and pos in this zone can be a false positive
ou false negative. So, it is better not to find the eplet only in gray zone. Farther
from the Cutoff is found na eplet, more positive or negative is the eplet.
Eplet report
In order to facilitate compatibility analyses, once the User clicks an eplet, a
report to it is generated and showed as seen in Figure 19. The eplet 32L was
found in NEG set, POS set and Gray zone set. Although it is an acceptable
mismatch, once it is both in NEG and POS sets, it can be a false negative
because this eplet is only in the negative part of the Gray zone. The eplet 62GE
is only in the POS set and, additionally, it was found in three molecules as the
possible unique positive eplet, including in the molecule B*5801 (very far from
cutoff). In this case, there is a high positivity associated with this eplet. The
70KHA is similar to 62GE, however it was found in the donor HLA typing. On
the other side, the 44RE has a high negativity associated with this eplet
because it was found in HLA molecules with normal values to far from cutoff.
Figure 19 – Three diferent eplets and their reports. The report gives a more accurate
vision about the selected eplet.
Histocompatiblity maps and searches in databases
The xHLA program can use the sets acceptable and unacceptable eplets to
perform searchers in databases in order to find the best/worst matching to a
donor or recipient. This facility can increase the reliability and the chances of
highly sensitizes patients in the waitlist to find an organ.
Exporting the eplets map
The eplets map can be exported to MS Excel easily, in order to facilitate the
medical reports. In this case, choose one of the eplets sets (Table 1) and click
the button
to export it.
xHLA and the LabSystem Gen Architeture
The xHLA tool uses the architecture provided by LabSystem Gen to perform its
tasks. The LabSystem Gen provides tools and a framework that allow
researchers to build a Laboratory Information Manager System (LIMS) without
programming efforts. The architecture of the System, as it is shown in Figure 20,
provides three layers: Application layer, Data access layer and Data storage
layer. On the top of the architecture is the Application layer (xHLA is in this
layer). The first step towards the creation of a LIMS is to design the Data and
Application models using the Designer tool. The Data Model includes a domainspecific data model describing the entities (tables) and relationships
(associations between tables) that the researcher wants to manage. The
application model, in turn, describes properties that allow the researcher to
customize the look and feel of the LIMS graphical user interface (GUI). As show
in Figure 20, on top, a Data and Application models can be created based on an
experimental plan using the modeling tools provided by Designer. In
laboratories. where data are stored in spreadsheets or legacy systems, an
automatic approach, using reverse engineering is also possible. In this process,
they are used as input to generate the Data model. However, the Application
model is still defined using the Designer modeling tools.
Figure 20 - LabSystem Gen’s general architecture.
In the middle layer is the eDA framework. The eDA framework provides
transparent access to data stored in databases and performs three major tasks
to Application layer: 1. It retrieves data stored in public HLA databases on web.
These data can be used in complex analysis performed by bioinformatics tools
developed in LIB; 2. It provides functions to store and to retrieve experimental
data stored in local relational database and; 3. It interprets a designer’s data
model and generates a relational schema in the local database (transformation).
This generated relational schema stores a copy of the Designer’s model for
future reference and the experimental data.
Public HLA databases and the local database are in the Data layer. The local
database is a backend MySQL relational database provided by System that can
be accessed by applications through the eDAframework. Experimental data can
be collected and managed using the Graphical User Interface (G.U.I.). The
desktop-based GUI is a generic program that dynamically build forms for users
manage data, based on the look and feel specified in the Application model.
The web-based GUI, is a set of web pages created dynamically by the Web
Server Application and, as desktop-based G.U.I. forms, they are based on
Application model and are used for users manage data.
xHLA, as shown in Figure 20, uses the eDA framework functionalities to manage
data in its local relational database and to access data existing in web
resources.
Report BUGS
Please, any bug report to [email protected]
Manual revision
Unfortunately this manual was not revised until 05/05/2010. Any comments
about errors and grammar, please help us to improve it reporting to us.
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