No species found from stool.fasta file in HUMAnN tutorial

Microbial community profiling HUMAnN
1

Hi!

In section 4.2 of the HUMAnN 3.0 tutorial, I had downloaded all 6 fasta files:
-rw-r–r-- 1 lnp524 staff 2.6M Apr 22 12:24 763577454-SRS014459-Stool.fasta
-rw-r–r-- 1 lnp524 staff 2.6M Apr 22 12:24 763577454-SRS014464-Anterior
nares.fasta
-rw-r–r-- 1 lnp524 staff 2.7M Apr 22 12:24 763577454-SRS014470-Tongue_dorsum.fasta
-rw-r–r-- 1 lnp524 staff 2.7M Apr 22 12:24 763577454-SRS014472-Buccal
mucosa.fasta
-rw-r–r-- 1 lnp524 staff 2.6M Apr 22 12:24 763577454-SRS014476-Supragingival
_plaque.fasta
-rw-r–r-- 1 lnp524 staff 2.7M Apr 22 12:25 763577454-SRS014494-Posterior
_fornix.fasta

Following the tutorial, when I ran this command on the first file (stool.fasta) I did not get any hits with either MetaPhlAn, bowtie2 or diamond:

humann -i 763577454-SRS014459-Stool.fasta -o hmp_subset

Creating output directory: /Users/lnp524/humann_dbs/HMP_downloads/hmp_subset
Output files will be written to: /Users/lnp524/humann_dbs/HMP_downloads/hmp_subset
Running metaphlan …
Total species selected from prescreen: 0
Selected species explain 0.00% of predicted community composition
No species were selected from the prescreen.
Because of this the custom ChocoPhlAn database is empty.
This will result in zero species-specific gene families and pathways.
Running diamond …
Aligning to reference database: uniref90_demo_prots_v201901b.dmnd
Total bugs after translated alignment: 0
Total gene families after translated alignment: 0
Unaligned reads after translated alignment: 100.0000000000 %
Computing gene families …
Computing pathways abundance and coverage …
Output files created:
/Users/lnp524/humann_dbs/HMP_downloads/hmp_subset/763577454-SRS014459-Stool_genefamilies.tsv
/Users/lnp524/humann_dbs/HMP_downloads/hmp_subset/763577454-SRS014459-Stool_pathabundance.tsv
/Users/lnp524/humann_dbs/HMP_downloads/hmp_subset/763577454-SRS014459-Stool_pathcoverage.tsv

The other fasta files worked just fine :slight_smile:

Maybe the database should be updated for a smooth tutorial?

Have a great day!
Caroline

2

Good note. I think the issue is that the tiny HUMAnN demo database doesn’t have all of the species present across all of these demo inputs. If you use the full HUMAnN database for this part of the demo you should get some output. The first part of the tutorial (working on the one demo.fastq sample) is designed to work with the tiny databases that come with the HUMAnN software.