MSigDB databases for mouse

[Mike]

Hi, There:

I went to the download page for MSigDB databases, and realized all databases files are for human if I am correct. Could someone let me know where I can get the mouse versions of the same MSigDB databases files (gmt etc)?

Thx a lot in advance!

Best

Mike

[Anthony Castanza]

Hi Mike,

The officially released versions of MSigDB only offers Human gene sets, however we provide mapping (.chip) files which perform orthology conversion for Mouse and Rat datasets so that data from these model organisms can be used with the Human MSigDB content. These annotation files are available through the GSEA interface or from our downloads page.

That said, lately we've started piloting an early draft of a series of collections that are natively mouse and don't require orthology conversion for mouse datasets, this early draft is available from here: https://www.gsea-msigdb.org/gsea/msigdb/mouse_geneset_resources.jsp and is currently version-consistent with MSigDB 7.4 (i.e. the versions of GO, Reactome, Ensembl etc used to build this draft of mouse resources is the same as the official Human version).

While these data are available, the Human MSigDB with mouse orthology chips remains the recommended way to run GSEA on mouse datasets.

-Anthony

Anthony S. Castanza, PhD

Curator, Molecular Signatures Database

Mesirov Lab, Department of Medicine

University of California, San Diego

[Mike]

Hi, Anthony:

Thx so much for the prompt response and very helpful information. 

Just want to make sure what you mentioned: this early draft is available from here: https://www.gsea-msigdb.org/gsea/msigdb/mouse_geneset_resources.jsp 

I did download the files and checked, For example for MH: mouse-ortholog hallmark gene sets, there are three files : 

mh.all.v0.2.entrez.gmt

mh.all.v0.2.symbols.gmt

mh.all.v0.2.metadata.txt

Looks like these two gmt files are the same as gmt files of human version hallmark gene sets that I used before, except that all gene IDs are converted to mouse IDs, and ready to use, is my understanding correct? What is the metadata.txt file for? 

Thx again and best

Mike

[Anthony Castanza]

Yes, that's correct, the GMTs are effectively equivalent. In the case of the Hallmarks we worked with the Bult lab at MGI to convert them from Human Gene Symbols to MGI Gene IDs and then used data from Ensemble's biomart to map the MGI IDs to Mouse Gene Symbols and NCBI Gene IDs. The mouse gene ID to Gene Symbol/NCBI Gene ID mapping is an equivalent process to what we use for mapping human gene IDs from various sources to Human Gene Symbols for MSigDB proper.

Because we don't have full gene set pages in the website for the mouse gene sets currently, which is what we use for sharing metadata like original publication sources, authors, contributors, links into pathway visualizations, we've compiled that metadata into the .metadata.txt files for distribution until that more complete support is implemented.

-Anthony

Anthony S. Castanza, PhD
Curator, Molecular Signatures Database
Mesirov Lab, Department of Medicine
University of California, San Diego

--
You received this message because you are subscribed to the Google Groups "gsea-help" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gsea-help+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gsea-help/fbae9849-aec3-4732-82ea-553f1761d4bcn%40googlegroups.com.

[Mike]

Cool! Thank you so much for your great help, Anthony! Have a great weekend! Mike