Hello,
I am a new user of the humann, and I was wondering if it is possible to obtain functional tables not normalized by RPKs, in order to play my self with certain methods.
If not, can I find the length of the genes in one of the files in order to revert the normalization ?
Many thanks for your help,
Alban.
I am sorry if my question wasn’t clear enough.
Is there anyone to discuss with?
Apologies for the delayed reply.
It’s not possible to output anything more raw than RPKs from native HUMAnN.
You can find the nucleotide-equivalent lengths of HUMAnN coding sequences attached at the end of their headers in the ChocoPhlAn database, the UniRef databases, and alignment reports against those databases (e.g. SAM output from pangenome mapping). Note however that HUMAnN uses a per-alignment effective length when computing RPKs, so it’s not as simple as multiplying RPKs by gene length in kilobases to compute raw hits.
I forgot to say thanks franzosa :).
Here it is
Alban