Plant CYP Catalytic Reaction Database

PCYPCatDB — a structured database of plant CYP catalytic reactions

This database is literature-evidence–centered. It integrates multi-agent large language model (LLM) extraction, context-aware species enrichment, and RDKit-based cheminformatics consistency validation to transform unstructured publications into machine-readable records—including CYP names, plant sources, protein sequences, substrate and product names, SMILES, evidence excerpts, and bibliographic provenance.

398 Species
1405 CYP enzymes
2490 Reactions

Monthly curation pipeline

On the 1st of each month, multi-agent LLM curators screen new PubMed literature, extract reaction records, and apply RDKit ΔMass validation. Release v2026.06.