CANYA

NewsecondsPaperCodeProduction Ready

Amyloid nucleation propensity prediction

Readiness

Readiness: Production Ready
Smoke: passed
Artifacts: passed
Production: production_ready_cpu_tensorflow
Storage: <5GB
Strategy: shared python312_slim.sif + tool ext3 overlay
Next: Optional: add batch FASTA upload support and expose ensemble/per-window columns more explicitly in the result view.

SOP validation used upstream CANYA revision 3a1efceca9f9b6eeece31664d630c0f3fca345a1 at /media/nik/seagate_nik/bio_server/tools/canya/source. Shared Python/TensorFlow env reused from /home/nik/micromamba-root/envs/canya-py39 via PYTHONPATH, avoiding dependency duplication. Standalone Slurm job 371 scored 367 example sequences with scores in 0-1 (min 0.082274, max 0.872746, mean 0.344159). Real runner API job d227a97120e8 submitted Slurm job 373 and completed with 10 downloadable artifacts including results.csv, canya_scores.txt, prediction.json, summary.json, stdout/stderr, and Slurm logs.