EquiDock

1-5 minutesPaperCodeProduction Ready

Rigid body protein-protein docking

Readiness

Readiness: Production Ready
Smoke: passed
Artifacts: passed
Production: production_ready_cpu_mode
Storage: <5
Strategy: compiled/binary SIF or shared python312_slim + ext3 overlay
Next: Use validated CPU DB5-case EquiDock adapter; add user-uploaded receptor/ligand mapping and GPU validation later if needed.

Production evidence 2026-07-01: source https://github.com/octavian-ganea/equidock_public revision ac2c754399bf20b50a27d86dbff4f6669788d47f installed at /media/nik/seagate_nik/bio_server/tools/equidock/equidock_public with micromamba env /home/nik/bio_server_venvs/equidock_py39. Worker Slurm benchmark 303 on cpu passed with wall 0:02.91 and output/GT sanity; API runner job remote ce0e5efb5b52, Slurm 309 on cpu completed. Runtime 2.827s. DB5 1AVX case produced 1AVX_l_b_EQUIDOCK.pdb, atom counts input/output 1286/1286, input-vs-GT RMSD 124.5493 A, output-vs-GT RMSD 35.5594 A. Downloadable artifacts include output PDB, results.csv, summary.json, harness, stdout/stderr and slurm logs. Caveat: upstream inference script is not CLI-parameterized and has hardcoded paths; adapter uses runtime source substitution without editing upstream. GPU not validated.