Humann2 unpack pathways result table format

Skip to first unread message

Angel Dai

Dec 3, 2019, 11:39:03 AM12/3/19
to HUMAnN Users
Hi dear developers,

I was wondering for the humann2_unpack_pathways script, 

1. Should I use cpm normalized table for both genefamily and pathway abundance table right?

2. How to interpretate the results? 

The command i ran is 
humann2_unpack_pathways --input-genes humann2_genefamilies_cpm.tsv --input-pathways humann2_pathabundance_cpm.tsv \
  --remove-taxonomy --output humann2_pwa_unpacked.tsv 

And a sample output table is as attached.

The header row still has all of the sample names, but the following rows generally is the "pathway|genefamily and then some number (I think its cpm)" format. 

I was wondering why we can't see each sample's cpm in that corresponding pathway|genefamily?

How is the number calculated from the two input tables?

Thank you very much for your time


Eric Franzosa

Dec 11, 2019, 4:00:44 PM12/11/19
This script won't change the units of the input genes. Rather it's showing you how different genes in your database were mapped to pathways during the pathway quantification step. This involves a two-step mapping from UniRef90/50 -> MetaCyc reaction and then MetaCyc reaction -> MetaCyc pathway. The row headers contain "Pathway|UniRef" pairs.


You received this message because you are subscribed to the Google Groups "HUMAnN Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to
To view this discussion on the web visit
Reply all
Reply to author
0 new messages