Hi, Im using HUMAnN4 for the gene profiling of soil samples with an average sequence depth of 7.5M reads. I used this script with uniref90 (humann -i ${name}_R1R2.fasta.gz --output Humann4output --protein-database ~/uniref --translated-subject-coverage-threshold 10 --threads 96 --bypass-prescreen) and it gave me a large table with 40000 genes, although ~90% of them were unclassified and from the classified ones, they didnt have any gene names even after relabelling them.
I wanted to download uniref50 but it says it`s not available and I could only choose uniref90. I appreciate hearing any thoughts about this issue.
Please take a look at the HUMAnN v4a release notes. To simplify things, we no longer require users to choose between UniRef90 and UniRef50. Instead, all search is done against subsets of UniRef90 (note that UniRef50 was also a subset of UniRef90) and translated search uses relaxed, UniRef50-like homology criteria.