How to do GSEA from gene profiles obtained from GEO

Question

[stewiegriffin]

I am new to bioinformatics, I want to code GSEA using gene profiles derived from the GEO database and using my custom gene set. How do I calculate the Enrichment score for each of the cohort in the GEO data set and get the p-values for distributions generated from k random permutations

[Santosh Anand]

Read the FAQ: http://software.broadinstitute.org/cancer/software/gsea/wiki/index.php/FAQ

The various scenario of running GSEA are clearly explained there.

Answer 1

[stewiegriffin]

Thanks for the link, GSEA could not rank the gene with only a single cohort.

How could I create a rank list with a data set having multiple profiles, but no control to compare with.

[Santosh Anand]

You must need one metric to rank the data. Because, the algorithm looks if there are enriched set at the top or bottom of a ranked list. Then you can run gseapreranked. How are your profiles created? Isn't there any metric which can rank the genes in the profile?

[stewiegriffin]

I am using the profiles from this data set( https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE22886) [HGU133A]. The expression levels of various probes are listed but can I use it to create a rank list.