MetaPhlAn Format Requirements for Input

I ran MetaPhlAn using --tax_lev s such that only species-level representations are written to the results file. Is that suitable to be input to HUMAnN? The HUMAnN 3.0 User Manual doesn’t really explain much apart from inputting the files via --taxonomic-profile bugs_list.tsv so it’s unclear what other requirements there might be for novice users to know of.

Sorry for the late reply - that should be OK. I believe we only parse out the species-level taxa anyhow for collecting appropriate pangenomes.

I have now run my analysis on the same input file without specifying a taxonomic profile and it works: “Total species selected from prescreen: 58”. So, --tax_lev s is a problem because 0 species are reported from the prescreen if it has been specified. @lauren.j.mciver, could you fix this or at least explicitly document it, if it’s supposed to happen, which I doubt it is?

HUMAnN does require the genus and species for taxonomy just to make sure we have the taxon correct. Sorry if this is not included in the documentation. I will make sure to add a note about it.

Thank you,
Lauren