Gene set enrichment analysis for genetic interactions

Question

[rubic]

Hi,

I have a list of pairs of genes that are simultaneously knocked out which have a corresponding score, for example an effect size and/or a corresponding p-value of the cell's growth compared to a control, which is better known as the result of a synthetic lethality screen.

I want to test which biological pathways are enriched at the top and bottom of that list. Unlike a traditional gene set enrichment analysis, in my case each entry in the list is a pair of genes rather than a single one, hence possibly mapping to two biological pathways. One obvious solution is to simply break down each pair and have them as separate, yet tied, entries, and GSEA that list. But I was wondering if anyone has come across something else.

[rpolicastro]

Do you have all (or most) pairwise combinations for a defined set of genes? If so you have a natural dataset for network analysis which might allow for more fruitful analysis.

[rubic]

No, these are pairs of paralogs, so by definition the genes are exclusive to a single pair.

[rubic]

Can please you tell more about the network analysis you've mentioned? Perhaps a link to a package/tutorial/papr? Thanks