DLKcat

secondsPaperCodeProduction Ready

Deep learning-based kcat prediction

Readiness

Readiness: Production Ready
Smoke: passed
Artifacts: passed
Production: production_ready
Storage: <5
Strategy: validated; reuse existing report/container on backend
Next: Monitor user-submitted DLKcat jobs; extend adapter to accept arbitrary uploaded TSV inputs beyond the upstream example path.

Production evidence 2026-07-01: source https://github.com/SysBioChalmers/DLKcat revision 7c15d0d4a7ac029f9d75564d9f2a93874aeaaec7 staged at /media/nik/seagate_nik/bio_server/tools/dlkcat. Runtime uses shared /media/nik/seagate_nik/bio_server/containers/shared/python312_slim.sif plus /media/nik/seagate_nik/bio_server/containers/dlkcat/dlkcat_py312_overlay.ext3. Worker Slurm job 321 matched upstream example output exactly. Real API runner job remote 15abf7ddacd8, Slurm 325 on cpu completed. Runtime 11.012s. Benchmark produced 3 kcat predictions with range 2.6097-57.4268 1/s; downloadable artifacts include input.tsv, dlkcat_output.tsv, results.tsv, results.csv, summary.json, stdout/stderr, and slurm logs.