Hi everyone,

I am working with metagenomic functional abundance data, specifically KO-level counts derived from metagenomes. I have performed differential abundance analysis using DESeq2 between two sample groups. After that, I would like to test whether KEGG pathways or KEGG modules show coordinated functional shifts between the groups.

My question is: is GSEA/fgsea statistically appropriate for metagenomic KO abundance data, assuming the input is a KO-level differential abundance table rather than transcriptomic gene expression data?

I do both GSEA and ORA analysis. So would you recommend GSEA/fgsea or ORA in this context or both?

Any advice would be greatly appreciated.

Thanks!