Progress Table

From Informatics

Jump to: navigation, search
Date Recorded Tasks Completed
10/2/2009 - half day Freed disk space on my desktop for caArray installation. Went over Matt's notes.
10/5/2009 - 10/6/2009 Created eclipse project for Matt's extensions. Have found that can compile Matt's code for 2.2.0 and 2.2.1 but can't compile for 2.3.0. Went again over Matt's notes.
10/13/2009 - 10/16/2009 caArray 2.2.0 installation.

UPT installation for caArray 2.2.0; getting some basic understanding of UPT, caArray and installation process, resolving db issues.

Created users for caArray2.2.0 (superadmin, caaray user admin, caaray user) associated user with group.

Created caaray application.

Logged into caArray instance as a carray user that was created in the previous step in UPT; created experiment, uploaded datasets.

Points from 1 to 7 on caArray 2.2.0 tasks list are done.

10/21/2009 - 10/23/2009 Followed Matt's instruction to insert his code into caArray.

Logged into caArray and created another experiment and loadded another dataset using the web interface.

Developed a client application to access Matt's code to add an experiment and dataset into caArray from the command line. Modified run script, classpath to include missing libs. Also changed constants in Matt's code : server, jndi port, user, pswd, array design, assay type; created dir with files to upload.

After running Matt’s code experiment was created in Work Queue, each file as a separate sample, status - draft, not a Public Experiment, can see it after login. Running Matt’s code second time created another experiment with the same name and same data but with a different experiment ID.

All has been done on my local machine including installation of caArray, UPT and MySQL

Steps 7, 11, and 12 were successful, see comments above.

Points from 8 to 13 on caArray 2.2.0 tasks list are done.

11/4/2009 - 11/6/2009, 11/9/2009 caArray 2.2.1

Steps 5, 9, and 10 were successful. All issues that I had were related to installation and being able to run 2.2.1 and 2.2.0 with the same UPT and MySQL instance. All has been done on my local machine including installation of caArray, UPT and MySQL

Points from 1 to 11 on caArray 2.2.1 tasks list are done.

01/27 half day KC forum, Topic "Caarray Help" about Matt's extension for 2.2.1, investigated the issue, posted message.
01/28,01/29, 02/01 Started testing Mat’s extension for Caarray 2.3.0.

Run out of hd space, cleaning hd after nearly each test. Some files were lost in a previous hd cleaning, spent time recreating testing environment. Followed Matt's instruction to update the EAR file. Followed Matt's instruction to update the caArray client jar. Setting classpath, more jar files were needed than for 2.2.1. When running tests got hibernate errors. Email correspondence with Matt as wasn't sure if it is code or classpath issue. 17 hours Matt's code, 4 hours hd issues

02/02 Modified classpath after Matt's suggestions.

Fixed mismatch between array design and assay type. Rerun tests. 7 hours Matt's code

02/04, 02/05 IMPORT_FAILED on the files that didn’t have problems for 2.2.1

Email correspondence with Matt as he never saw such messages when running his extension. Exchanged testing data with Matt: created experiment on nih with my data for Matt to test, and get Matt's data to test on my pc. Matt's data were too big, left import of Matt's array design files overnight, get message out of space, need more hd space. 10 hours Matt's code, 4 hours hd issues

02/10 Get a new version of Matt's code; modified this new version for my tests; Rerun tests. Tests on my data passed.

7 hours Matt's code

02/12 MAGE-TAB data testing 1: issues with idf file, validation failed error, but detailed error description available only through the web interface.

MAGE-TAB testing 2: Zhong modified idf file, test passed. 7 hours Matt's code

Conclusion: code works as advertized; having better error checking and reporting will be very beneficial.

03/09 - 03/11 Correction for README.txt file: in paragraph 3.2 of the instructions, in two places "update_client.sh" should be replaced with "update_client_jar.sh".
  1. Went through Matt's instructions and my notes.
  2. in addition to Matt's instructions modified client code: in the file LoadArrayDataFiles.java changed: server host name,user name,user password, experiment title, assay type,array design,location of datafiles.
    1. The changes to class const; note that user name and password are not real and need to be replaced:
      private static final String SERVER_HOST = "MAGNET13";
      private static final int SERVER_JNDI_PORT = 31099;
      private static final String USERNAME = "yourUsername";
      private static final String PASSWORD = "yourPswd";
      public static String TITLE = "Test Affymetrix Mapping";
      public static String PROVIDER = "Affymetrix";
      public static String ASSAY_TYPE = "Gene Expression";
      public static String[] ARRAY_DESIGNS = new String[{"HG-U133_Plus_2" };
    2. The changes to the function main, note that Matt uses a relative path and i use an absolute path.:
      client.uploadFiles(project, "C:/data/caarray/dataFiles");
  3. In "C:/data/caarray/dataFiles" directory I have files: GSM183432.CEL GSM183432.CHP
  4. Created Eclipse project.
    1. to create a project in Eclipse:
      1. on the main menu bar, click File > New -> Java Project
      2. select "Create Project from existing source" radio button
      3. Click "Browse" button and navigate to caarray-cri-api-extension-1.0 directory, which is described in paragraph 1. of the README.txt file.
      4. Finish project creation.
      5. Project creation guidelines are in Eclipse Help, here we want to highlight only what is specific for our project.
    2. To set classpath for the project:
      1. Navigate to lib/java directory of caArray legacy client distribution that is described in paragraph 3.1 of README.txt file, and add all .jar files from this directory to the classpath.
      2. Add caarray-client-legacy.jar file, after it was modified as described in paragraph 3.2 of README.txt file, to the classpath.
    3. This files should be excluded from the project's Build Path: ProjectMgmtService.java, ProjectMgmtServiceBean.java
  5. Run the sample client program.
    1. To run in Eclipse:
      1. in outline view open LoadArrayDataFiles.java file.
      2. right click on the function main -> Run As -> Java Application.
  6. To verify that data are uploaded correctly go to caArray web interface, in my case it will be http://magnet13:38080/caarray/home.action and:
    1. login with your user name and password.
    2. Go to "experiment management"->"my experiment workspace", which should be on the Left Vertical Navigation Task Menu.
    3. You well see all your experiments. Look for experiment with the title that you have created and click on it.
    4. Click on the "Data" tab -> "Imported Data" tab, look at the column status, all files should have status imported and check boxes should be checked.
    5. Click on the "Download Data" tab, expand plus sign for each file, click on "Launch Download Job" tab, pop-up dialog window will appear, select save file radio button, click "OK" button, select location on your hd.
    6. File will be saved in .zip format, unzip files, there should be all the files that have been uploaded.
  7. Run the sample client program with MAGE-TAB data files.
    1. For this step i have changed the constant TITLE to:
      public static String TITLE = "Test Affymetrix Mapping MAGE";
    2. and have changed the line in function main:
      client.uploadFiles(project, "C:/data/caarray/MAGEdataFiles");
    3. in "C:/data/caarray/MAGEdataFiles" directory I have files: GSM183432.CEL GSM183436.CHP GSM183441.CEL GSM183432.CHP GSM183437.CEL GSM183441.CHP GSM183433.CEL GSM183437.CHP GSM183442.CEL GSM183433.CHP GSM183438.CEL GSM183442.CHP GSM183434.CEL GSM183438.CHP GSM183443.CEL GSM183434.CHP GSM183439.CEL GSM183443.CHP GSM183435.CEL GSM183439.CHP vivanco_PTEN_affy_disease_state.idf GSM183435.CHP GSM183440.CEL vivanco_PTEN_affy_disease_state.sdrf GSM183436.CEL GSM183440.CHP
  8. Code was tested with caArray version 2.3.0.1
  9. Wrote these instructions.
03/17 worked on caarray-importer-1.0.2
  1. Went through Matt's instructions and my notes.
  2. In addition to Matt's instructions modified client code; changes in the file CaArrayImporterExample.java: server host name,user name,user password, experiment title, assay type,array design,location of datafiles.
    1. The changes to class const; note that user name and password are not real and need to be replaced:
      private static final String SERVER_HOST = "MAGNET13";
      private static final int SERVER_JNDI_PORT = 31099;
      private static final String USERNAME = "yourUsername";
      private static final String PASSWORD = "yourPswd";
      public static String TITLE = "Test Affymetrix Mapping";
      public static String PROVIDER = "Affymetrix";
      public static String ASSAY_TYPE = "Gene Expression";
      public static String[] ARRAY_DESIGNS = new String[{"HG-U133_Plus_2" };
    2. The changes to the function main, note that Matt uses a relative path and i use an absolute path.:
      client.uploadFiles(project, "C:/data/caarray/MAGEdataFiles");
  3. build.properties file has been changed too, in my case property server.base.dir has been changed and build.properties file looked like this:
    client.legacy.zip=caarray-client-legacy.zip
    server.base.dir=C:/programs/caarray/caarray2_3_0
    server.deploy.dir=${server.base.dir}/jboss-4.0.5.GA/server/default/deploy
    server.ear=${server.deploy.dir}/caarray.ear
  4. From command line run several ant targets according to instructions in README.txt.
  5. Verified that data are uploaded correctly as described in the previous entry.
04/13 - 04/16 worked on CSC(caArray Standalone Client)
  1. Found a bug in the first version of the code; reported the bug. Next day was advised to download the fix. Downloaded the second version of the code, bug was still there; reported. Was advised how to modify the code; fixed the code accordingly. After many attempts was able to run the application successfully.
  2. Observations:
    1. General observations - Application comes with eclipse project configured for their particular environment, and doesn't work out of the box. Also, it would be very beneficial to provide an ant script in order to run the application from the command line.
    2. Observations from running the application - sometimes the application became unresponsive and I had to kill the process. Exceptions that I was getting were not very informative. And sometimes application ended without any messages and no data were uploaded, results were checked through the web interface. Wasn't able to find a way to save input data and each application run had required entering input data, which were mostly the same.
05/14 Worked on CSC(caArray Standalone Client)
  1. There were great improvements in usability. Issues that were raised during the previous round of testing were fixed:
    1. There is an ant script now and the user don't have to configure an eclipse project.
  2. The user can ran the script out of the box.
    1. Most of the input parameters that are entered on the GUI fields are saved in a properties file and the user don't have to retype them.
    2. GUI isn't entering unresponsive state anymore, and there is no need to kill the application.
  3. Tests were repeated several times, results were tested through the web interface.
    1. First, was trying to upload files with an experiment name provided but didn't notice any new data files in this experiment.
    2. Second, was trying to upload files without experiment name provided but new experiments wasn't created.
  4. There was no problem importing the same data files through web interface.
Personal tools