compchem-molpal-screening

What problem does it solve?

Active-learning virtual screening workflow for aquatic ecotoxicology, enabling rapid prioritization of toxicants by LC50 using a molpal Bayesian optimization loop.

Core Features & Use Cases

• Active-learning virtual screening: Uses molpal to prioritize seed compounds by acute aquatic toxicity (LC50) within a Bayesian optimization framework.
• Validation & data quality: Enforces RDKit SMILES parsing, separators, and mechanism_of_toxicity length to ensure reliable downstream modeling (EPA Category I).
• Use Case: Researchers can plug in seed_compounds.csv and obtain a prioritized, validated pool for screening in computational ecotoxicology workflows.

Quick Start

Provide initial seed compounds and run the molpal-based screening to generate a ranked pool of compounds by LC50-derived scores.