molclaw-karmadock-tool

KarmaDock automates protein-ligand virtual screening by running batch docking on ligand SMILES libraries against a target protein PDB, generating ranked pose SDF files and summary metrics for downstream analysis without manual orchestration.

Protein-ligand molecular screening with KarmaDock produces three scoring metrics: karma_score, karma_score_ff (force-field corrected), and karma_score_aligned (alignment-corrected). These robust scores ensure reliable ligand ranking and rescoring for high-throughput campaigns.

A crystal ligand MOL2 file is optional for pocket localization in KarmaDock virtual screening. You must provide a target protein PDB, but the crystal ligand MOL2 is only supplied when you need precise pocket definition for batch ligand docking workflows.

KarmaDock supports configurable score_threshold, batch_size, and random_seed parameters for reproducible ligand ranking. These operational modes allow precise control over high-throughput virtual screening campaigns and ensure consistent results across batch docking runs.

Retrieving ranked SDF pose files requires completing the mandatory output download step after KarmaDock finishes virtual screening. This enforced policy ensures you locally download exported pose files for verification, reproducibility, and downstream analysis.

Validating a large compound library before rescoring uses KarmaDock's dry-run validation mode. You can pre-screen a 100k-compound library with configurable batch sizes to obtain top-ranked poses efficiently for subsequent MD validation or interaction fingerprint analysis.