skip normalization

[jorondo1]

Future modification (not a priority) :

Remove humann_renorm step; keep raw counts for downstream analysis to account for the compositional nature of data. All subsequent humann steps can be performed on raw counts.

[jflucier]

do you mean this line

[jflucier]

must be it since only place humann_renorm is called.

Please confirm and i will delete & release.

[jorondo1]

exactly, so the loop right above it ($norm_method == "cpm") becomes irrelevant and the subsequent files will not have the $norm_method suffix

while you're at it, the operation at this line can actually be done using a humann utility, namely this one. It takes the .tsv created by the humann_regroup script loop and splits them exactly like I want (stratified levels + community level, called "unstratified" here)

[jorondo1]

did you get a chance to do that ?

must be it since only place humann_renorm is called.

Please confirm and i will delete & release.

[jflucier]

commented this step. See here

released on ip29

[jflucier]

damm just saw we need to remove norm for loop, will proceed with this modif

[jflucier]

removed the for loop.

please validate here

[jorondo1]

yes j'allais justement t'écrire pour dire que faut enlever la loop ça a l'air good