Hi, 

What software are you using for your pathway analysis? I regularly use IPA or DAVID and I use a cut-off of FDR < 0.05 to rank the pathways based on FDR and concentrate on the ones that are highly significant but also have a relevant biological meaning. As far as I know there's no fold enrichment/FDR cut-off standard. 

M

I use some of the same software as Matina and agree with her approach. DAVID enrichment p-values do include some adjustment for pathways with different numbers of genes (although, of course, it is not perfect). But you might want to check out the linked article (Huang et al, 2009) describing DAVID's approach, and the interpretation of results, especially the two paragraphs under "Remaining Questions and Challenges in the Field: 1. Realistically positioning the role of enrichment P-values..." 

On a more philosophical note, 'biologically meaningful' fold enrichment cutoff makes it sound like

• we can tell that a pathway change is biologically meaningful based on overrepresentation analysis
• we know what the fold enrichment (or enrichment P-value) cutoff should be to isolate biologically meaningful results
• a single fold enrichment cutoff would apply to all of the thousands of pathways

IMHO, results from an over-representation analysis can't determine a pathway's biological meaningfulness, but candidate pathways can be flagged for further study, and this analysis approach can be used to determine the degree to which a given pathway responds to an intervention. Even a strongly over-represented pathway could be epiphenomenal, and a weak pathway could have a keystone biomolecule with broad influence. Intervening in the pathway, and showing that such an intervention had an effect on the system- that is what would prove biological meaningfulness.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2615629/