Hi All,
I'm new to these forums, but hoping you can help me with a question.
I am using phyper in R to compute p values that assess the significance of overlapping gene sets, though from whole exome sequencing results, not gene expression data.
1-phyper((#overlaps-1), disease list size, (total # genes-disease list size), PPI list size) example: 1-phyper((9-1), 353, 20000-353, 182)
where my PPI list size is the # of genes in my own dataset and disease list size is the number of genes in the gene set of interest.
I have performed a Bonferroni Correction to correct for multiple comparisons/tests, but feel like the proper way to assess significance is using a permutation test. I am wondering whether folks have good recommendations of R packages or scripts that will perform this kind of test WITHOUT gene expression data, and in general feedback about how to perform the permutation test. I have little statistical knowledge and little experience using R.
Thanks for any help you can give me!
If you need to correct for multiple testing and find Bonferroni too conservative, the standard alternative is to control the false discovery rate, typically using the Benjamini-Hochberg procedure (available among others in the p.adjust function in R). See here.