GSEA v4.0.x Release Notes

From GeneSetEnrichmentAnalysisWiki

(Difference between revisions)
Jump to: navigation, search
Line 1: Line 1:
-
[http://www.broadinstitute.org/gsea/ GSEA Home] |
+
61[http://www.broadinstitute.org/gsea/ GSEA Home] |
[http://www.broadinstitute.org/gsea/downloads.jsp Downloads] |  
[http://www.broadinstitute.org/gsea/downloads.jsp Downloads] |  
[http://www.broadinstitute.org/gsea/msigdb/ Molecular Signatures Database] |  
[http://www.broadinstitute.org/gsea/msigdb/ Molecular Signatures Database] |  
Line 7: Line 7:
<br />
<br />
<h2> GSEA Desktop v4.0.0 (Aug 2019)</h2>
<h2> GSEA Desktop v4.0.0 (Aug 2019)</h2>
-
<p>
+
 
-
<h3>Notes</h3>
+
-
  <ul style="margin-top:0pt;margin-bottom:0pt;">
+
-
    <li dir="ltr" style="list-style-type:disc>The algorithm in this version of GSEA is unchanged from the former code line and produces results equivalent to the previous version (v3.0).</li>
+
-
    <li dir="ltr" style="list-style-type:disc>GSEA Desktop v4.0.0 requires either Java 11 or Java 8.  Java 11 is preferred; we provide bundles with an embedded OpenJDK 11 JVM for Mac, Windows (64-bit), and Linux.</li>
+
-
  </ul>
+
-
</p>
+
-
<p>
+
<h3>Improvements and Bug Fixes</h3>
<h3>Improvements and Bug Fixes</h3>
-
  <ul style="margin-top:0pt;margin-bottom:0pt;">
+
* Updates for the MSigDB v7.0 release.
-
    <li dir="ltr" style="list-style-type:disc>Updated for the MSigDB v7.0 release.</li>
+
* The code has been updated to be compatible with Java 11 and the (free) OpenJDK version of Java.
-
    <li dir="ltr" style="list-style-type:disc>Migrated code for Java 11 compatibility and corresponding better security.</li>
+
* New options for downloading and launching the application:  
-
    <li dir="ltr" style="list-style-type:disc>Performance improvements: we are seeing a roughly 40% improvement in our informal testing.</li>
+
** Windows installer, Mac app, and Linux bundle, each with an '''embedded OpenJDK Java 11''', which means no separate Java installation is needed.   
-
    <li dir="ltr" style="list-style-type:disc>Created dedicated Windows installer, Mac.app, and Linux bundle, each with an embedded OpenJDK 11 to avoid issues with the user needing to install Java.  Created new command-line scripts for launching the GUI and running batch analyses so the user can avoid needing to deal directly with the more complex Java 11 command-line.</li>
+
** Command-line scripts for running batch analyses, which means users no longer need to enter Java commands directly. Note that a '''separate Java 11''' installation is required.
-
    <li dir="ltr" style="list-style-type:disc>Removed implicit dependence on out-of-date internal annotation CHIP file.  Removed other special CHIP handling code to avoid conflicting with the results of the new symbol conversion from MSigDB 7.0 and better support RNA-Seq workflows. Annotations now come from the CHIP used to collapse, or else from the dataset itself (and are left untouched).</li>
+
** Note: Launching the application using Java Web Start still requires a '''separate Java 8''' installation.
-
    <li dir="ltr" style="list-style-type:disc>Changed the Cytoscape integration to avoid issues with platform-specific launching issues, newer Cytoscape versions, multiple installations, custom installations, and so on.  The integration now prompts the user to launch Cytoscape manually before it attempts to connect.  Cytoscape 3.3+ is required.</li>
+
* Miscellaneous performance improvements and code simplifications. These add up to a roughly 40% improvement in our informal testing.
-
    <li dir="ltr" style="list-style-type:disc>Switched from Ant to Gradle for builds.</li>
+
* Changed the Cytoscape integration to avoid issues with platform-specific launching issues, newer Cytoscape versions, multiple installations, custom installations, and so on.  The integration now prompts the user to launch Cytoscape manually before it attempts to connect.  Cytoscape 3.3+ is required.
-
    <li dir="ltr" style="list-style-type:disc>Many code simplifications and removal of unnecessary code.</li>
+
* Fixed an indefinite wait cursor in the Leading Edge Tool.
-
    <li dir="ltr" style="list-style-type:disc>Fixed an indefinite wait cursor in the Leading Edge Tool.</li>
+
 
-
  </ul>
+
 
-
</p>
+
* DAVID I WANT TO REMOVE THIS ONE. UNLESS THERE ARE ANY USER-FACING IMPLICATIONS OF THIS CHANGE? IN WHICH CASE CAN THEY BE DESCRIBED INSTEAD? Removed implicit dependence on out-of-date internal annotation CHIP file.  Removed other special CHIP handling code to avoid conflicting with the results of the new symbol conversion from MSigDB 7.0 and better support RNA-Seq workflows.  Annotations now come from the CHIP used to collapse, or else from the dataset itself (and are left untouched).

Revision as of 20:19, 13 August 2019

61GSEA Home | Downloads | Molecular Signatures Database | Documentation | Contact


GSEA Desktop v4.0.0 (Aug 2019)

Improvements and Bug Fixes

  • Updates for the MSigDB v7.0 release.
  • The code has been updated to be compatible with Java 11 and the (free) OpenJDK version of Java.
  • New options for downloading and launching the application:
    • Windows installer, Mac app, and Linux bundle, each with an embedded OpenJDK Java 11, which means no separate Java installation is needed.
    • Command-line scripts for running batch analyses, which means users no longer need to enter Java commands directly. Note that a separate Java 11 installation is required.
    • Note: Launching the application using Java Web Start still requires a separate Java 8 installation.
  • Miscellaneous performance improvements and code simplifications. These add up to a roughly 40% improvement in our informal testing.
  • Changed the Cytoscape integration to avoid issues with platform-specific launching issues, newer Cytoscape versions, multiple installations, custom installations, and so on. The integration now prompts the user to launch Cytoscape manually before it attempts to connect. Cytoscape 3.3+ is required.
  • Fixed an indefinite wait cursor in the Leading Edge Tool.


  • DAVID I WANT TO REMOVE THIS ONE. UNLESS THERE ARE ANY USER-FACING IMPLICATIONS OF THIS CHANGE? IN WHICH CASE CAN THEY BE DESCRIBED INSTEAD? Removed implicit dependence on out-of-date internal annotation CHIP file. Removed other special CHIP handling code to avoid conflicting with the results of the new symbol conversion from MSigDB 7.0 and better support RNA-Seq workflows. Annotations now come from the CHIP used to collapse, or else from the dataset itself (and are left untouched).
Personal tools