Transcription Factor Enrichment

[Luís Eduardo Gonçalves]

Hi Everyone!

Recently I ran enrichment analysis based on a gene of my interest for correlation analysis. I was thinking if I can run this analysis regarding the three TFTs readings available on GSEA. If it is possible, what is the correct permutation type for this analysis?

Thanks

Luis

[Anthony Castanza]

Hi Luis,

There are two Transcription factor target collections, c3.tft.gtrd.v7.4.symbols.gmt subcollection, and the c3.tft.tft_legacy.v7.4.symbols.gmt sub-collection, these both describe predicted targets of a given transcription factor as determined by different methodologies. They should generally be run individually and we recommend against using the combined " c3.tft.v7.4.symbols.gmt" collection. The difference in the methodologies is described on the gene set details page: http://www.gsea-msigdb.org/gsea/msigdb/collection_details.jsp#TFT

As to the correct permutation type, this is going to depend on the number of samples you have in your dataset, generally I would consider 7 total samples to be the minimum to run phenotype permutation in the correlation mode, the nature of the correlation calculation has an advantage over the per-phenotype mode in this sense which generally requires 7 per group. If you don't have this minimum number of samples I gene_set permutation would be advised.

-Anthony

Anthony S. Castanza, PhD

Curator, Molecular Signatures Database

Mesirov Lab, Department of Medicine

University of California, San Diego