molclaw-karmadock-tool

What problem does it solve?

Automates large-scale protein-ligand virtual screening to produce ranked ligand poses and summary metrics, eliminating manual orchestration of docking, scoring corrections, and output collection for downstream analysis and verification.

Core Features & Use Cases

• Batch Virtual Screening: Run KarmaDock on ligand SMILES libraries against a target protein PDB with optional crystal ligand MOL2 pocket localization.
• Robust Scoring & Correction: Produces karma_score plus force-field and alignment-corrected scores (karma_score_ff, karma_score_aligned) for reliable ranking.
• Pose Export & Reproducibility: Exports ranked pose SDF files and enforces mandatory download policies to ensure local verification and reproducibility.
• Operational Modes: Supports dry-run validation, configurable score_threshold, batch_size, and random_seed for reproducible campaigns.
• Use Case: Pre-screen a 100k-compound library to obtain top-ranked poses for rescoring, MD validation, or interaction fingerprint analysis.

Quick Start

Run KarmaDock to screen a ligand library against a protein structure, generate ranked pose SDFs, and download the outputs for downstream analysis.