MeV v4.4 Released

The MeV development team is proud to announce the release of MeV v4.4.
It is now available for download at the TM4 website,

Please visit the TM4 Forums to get help or suggest improvements to the

Please consider filling out the MeV Survey. Doing so will let us know the things you do and don't like about MeV.

MeV v4.4 release
May 28, 2009

Major additions

BN Predict Cytoscape Plugin as MeV Extension
As an extension to the already published BN method, we have developed a Cytoscape plug-in that does predictive modeling of the BN described above. For each network loaded with an associated conditional probability table (CPT), the plugin calculates probabilities of expression (bin/state probabilities) for each network node. Documentation can be found at

MAGE-TAB File Loader
The MeV MAGE-TAB file loader allows a user to select and load expression data from MAGE-TAB (version 1.0) formatted files. The MAGE-TAB specification can be found at This format includes files that describe an experimental series, the samples used in the experimental series, optionally the array design of the microarray chips employed in the study, and a data matrix file containing gene expression data values.

New Cluster Manager Tool
The new Cluster Manager allows users to create, visualize, and manipulate clusters and includes numerous new features and enhancements to the previously existing clustering system.

RP Two-Class Unpaired and Two-Class Paired experimental designs.
MeV now supports the use of the Rank Products module for two-class experimental designs. Previously, only one-class functionality existed for this method. As with most other modules, the Cluster Selector feature is compatible with this addition.

Automated Linking to Online Databases
MeV can now automatically link to online databases, such as NCBI, AmiGO and others. When gene annotation is loaded using MeV's new annotation file loader (under the "annotation" subsection of the file loader) links will automatically be generated from some of these annotations to the appropriate online resource. The specific resources used can be customized for individual users, and expanded to other annotation types not loaded by the MeV automated annotation loader. See the MeV manual for more details.

Annotation Support for New Arrays
The MeV team has developed a new tool for creating the annotation files everyone depends on for their data analysis. We will be providing support for many new arrays in the future. These arrays will be available from the ftp site as well as via direct download from within MeV's file loaders. We have also generated new EASE datafiles that provide better, more complete EASE results.

Java WebStart Improvements
MeV can now accept a wider range of command-line arguments, improving its usefulness as a webstarted application. Instructions for getting started with webstart are available at

Assignment File Saving Synchronization
MeV's assignment file system has been synchronized between virtually all modules and experimental designs. Files saved in each module are compatible with the loaders of other modules. Additionally, the files themselves are saved in a human-readable format that can be opened, read and edited by users. See the MeV manual for more information.

Minor Additions

* The entire dataset is now displayed in each of the standard viewers - heatmap, expression graph, centroid and table view - immediately upon loading.
* New Proportional branch length feature added to Hierarchical Clustering.
* A new p-value viewer has been added to GSEA.
* Node rotation feature added to Hierarchical Clustering.
* Sample reordering feature added to heatmap viewers.
* Five standard viewers added to ?Original Data? immediately after data is loaded.
* Cluster Selection feature added to SAM module for "One-Class", "Two-class unpaired", and "Multi-class".
* New feature allowing multiple clustering by binning.
* A new interface to the annotation downloader makes annotation loading easier.
* Excel-style copying of data from result tables is now available.
* Behind-the-scenes work on state-saving makes it much faster and more robust.

* Fixed missing columns in One-Condition BETR output.
* Improved handling of missing data in BETR analysis.
* Tweaks to make MeV behave better without an internet connection.
* SAM no longer tags genes with zero variance as "Negative Significant".

Known Issues
* Results returned by the RP module in two-class unpaired analysis can become uninformative for datasets with greater than about 40 samples. When the number of samples is too high, the module will report RP values of zero for many of the genes. We recommend using the SAM module instead of RP for these datasets, as RP was designed to work best with small numbers of samples. We plan to have this issue addressed for the next MeV release.

Questions? Comments?
Please let us know in the MeV forums.

Posted by Eleanor Howe 2009-06-02

Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

No, thanks