Differential abundance Metaphlan

Hi there,

I’m wondering what is the recommended way to perform a differential abundance analysis on metaphlan results.

Let’s say we are working at the species level. Right now I’m considering two options :

  1. Perform Kruskal-Wallis tests on the raw metaphlan results + BH correction for multiple testing.
  2. Transform metaphlan results (proportion) into “absolute” abundance using number of reads per sample as a proxy, and feed these results to any existing DA method (e.g. DESeq2)

Thanks for your suggestions,
Erwan