GitHub
10/14/2026, 5:02 AMREADME.md
• If I modified data: I checked whether the root README.md is still up-to-date
• I made sure to submit the smallest dataset possible (e.g. try .tar.gz for larger directories)
Description
Test data for the rundbcan modules, whose tests now download the full 7.4 GB dbCAN database and whose query proteome gets no dbCAN HMM hits and no CAZyme gene clusters.
Added under `data/genomics/prokaryotes/bacteroides_fragilis/genome/`:
• region_2500000-3000000.faa.gz (90 KB): the 356 proteins of the 2.5–3.0 Mb region of the existing genome.gff.gz, translated from genome.fna.gz.
• region_2500000-3000000.gff (216 KB): the GFF lines for the same region, uncompressed because run_dbcan cannot read a gzipped GFF.
• db/dbcan.tar.gz (7.5 MB): a mini dbCAN database cut from the full one.
On the region, the mini database gives the same CAZymes, HMM hits, 14 gene clusters and 9 substrate predictions as the full database, in 18 s instead of 10.6 min.
TCDB, sulfatase and STP get 1–2 extra borderline hits, since E-values are smaller against a smaller database.
The README describes how each file was made.
The README also described the existing B. fragilis genome files as NC_006347 (strain YCH46); they are strain FDAARGOS_1225 (NZ_CP069563.1), so those lines are corrected.
🤖 Generated with Claude Code
https://claude.ai/code/session_01LydWahEnWZexTULEub5bfX
nf-core/test-datasetsGitHub
10/14/2026, 6:21 AMGitHub
10/14/2026, 6:34 AM