GitHub
10/28/2026, 6:53 PMcolabfold/colabfold module (nf-core/modules#13066), requested in review: Trp-cage (PDB 1L2Y, 20 residues) so a GPU prediction takes seconds.
| File | Size | Used for |
| ---------------------------------------------- | ----- | ------------------------------------------------------------------------------------------------------------------------ |
| data/proteomics/pdb/1l2y.pdb | 25 KB | templates input (--custom-template-path); model 1 of 1L2Y, header records kept |
| data/proteomics/colabfold/trp_cage.a3m | 31 B | query as a single-sequence A3M, so --templates runs without the ColabFold MSA server |
| data/proteomics/colabfold/trp_cage_pdb_hits.m8 | 52 B | pdb_hits input (--pdb-hit-file), one hit to 1l2y_A |
| data/proteomics/colabfold/local_pdb.tar.gz | 6 KB | local_pdb input (--local-pdb-path), one-entry mirror local_pdb/1l2y.cif made with ColabFold 1.5.5's convert_pdb_to_mmcif |
The README of data/proteomics describes the files and how they were made. All three GPU tests of the module pass locally with these files (prediction finds 1l2y_A as template in both template tests).
🤖 Generated with Claude Code
nf-core/test-datasetsGitHub
10/30/2026, 7:51 AMGitHub
10/30/2026, 7:52 AM