the order genes in the gene list on the topGO

Question

[azadeh.taati]

hello there

I work with topGO for the analysis of gene enrichment. I discovered that if the order of genes changes in the gene list, the results of p-values also change. I don't know how to order the genes in my gene list to make the results logical.

[dthorbur]

How big are these changes? I've not noticed that before, so I would run a quick permutation test. Randomly order your gene list using the sample function, and run the test 100-1000 times to quantify how often you lose significance around your results.

Otherwise, random input order seems like the most fair way to me.

I would also maybe check out if this gene order issue is found in other tools, like g:profiler2 for example.