Difficulty downloading the Humann database

[jsweagley]

The --updatehumann command is not working, it appears the link has been broken.

wget http://huttenhower.sph.harvard.edu/humann_data/uniref/uniref90/uniref90_protein_201901.tar.gz

and

wget https://g-227ca.190ebd.75bc.data.globus.org/humann_data/uniref/uniref90/uniref90_protein_201901.tar.gz

as per (https://forum.biobakery.org/t/humann-database-download/6690/17)

both fail to download with with an ERROR 404: Not Found.

The current version of Humann appears to have a new download link. Unsure if this will work appropriately with the version of the software included with TOFU-MAaPO.

http://huttenhower.sph.harvard.edu/humann_data/uniprot/uniref_ec_filtered/uniref90_annotated_v4_alpha_ec_filtered.tar.gz

[eikematthias]

Hi! Thanks for bringing this to my attention.

--updatehumann relies on the HUMAnNs database utility script. IN the version currently implemented in TOFU-MAaPO, the database is downloaded from http://huttenhower.sph.harvard.edu/humann_data/uniprot/uniref_annotated/uniref90_annotated_v201901b_full.tar.gz

At the moment this file is reachable from my side. That said, I have also seen occasional short outages in the past.

Could you please try running TOFU-MAaPO again? If it still fails, please attach the log file .nextflow.log, so I can pinpoint if this a problem of the pipeline or something local. Also, please check that the machine and containers you are running on have internet access and no proxy or filewall rules blocking the download.