Dear community,
in order to find significant enrichment in certain gene sets for my microarray data (2 conditions), I have tried out with different methods to calculate the gene scores, yielding different results in the number of enriched gene sets. In this way, the methods with the decreasing number of enriched gene sets (q-value < 0.05) are, for KEGG pathways: KS (81) > U-statistic (32) > FC (14) > t-statistic (0). The U-statistic 32 gene sets include the FC 14 gene sets, and the KS-statistics 81 gene sets include in turn the U-statistics ones. The intersection of KS, U and FC results equals FC. It might be possible KS-statistics yields many false positives.
I would be very very grateful if somebody could give me a piece of advice on this, since I do not know which method should be the most appropiate.
I am using gage package in R by the way.
Thanks
Different scores do different things. Try to understand how they relate to the question you're interested in e.g. if you're most interested in effect size then the fold change is what you should use but if you're more interested in statistical significance then look for one of the statistics taking into consideration the assumptions they make e.g. t-test assumes normally distributed data whereas KS doesn't.