molclaw-equiscore-tool

EquiScore batch screening evaluates ligand-pocket pairs using GPU-enabled inference to produce per-pocket prediction CSVs. It generates summary statistics for ranking, enabling rapid enrichment-style filtering of virtual screening results.

The end-to-end pipeline runs pocket extraction then EquiScore screening in one click. It returns organized run directories and ranked prediction CSV paths, making it suitable for hit prioritization and downstream validation workflows.

EquiScore supports GPU batch inference and multi-pose evaluation to accelerate scoring. It also includes dry-run validation features to verify inputs before executing the full virtual screening ranking process.

The automated workflow extracts binding pockets from docking outputs and scores ligand-pocket pairs using EquiScore. It produces ranked prediction CSVs from your SDF and PDB inputs, eliminating manual preprocessing for candidate prioritization.