Lefse Input file for functional data-TPM or raw counts?

Hi, Thank you for a great tool! I am interested in using Lefse on shotgun-sequenced data. I have come across some posts and queries regarding Lefse input file format for OTU tables, but can’t seem to find information on shotgun-sequenced functional data. Should the input file be raw counts, TPM or other normalized versions of it?


Hi Jigyasa,
If you’re still in the planning stages, my strong recommendation is to use our newer tool, MaAsLin 2, for feature-based testing (in general, and on shotgun-sequenced data specifically). It can perform a few basic transformations/normalizations internally to the program, and can also take in ready-to-go data of your choice.