ToxinDB

ToxinDB predicts biotransformation products of biogenic toxins and aggregates curated toxin, reaction, and metabolite data—including a global chemical space of roughly 550,000 toxins and putative metabolites—to support computational and experimental study of toxin metabolism.

Key Features:

• Extensive Database: Contains multifaceted data on more than 4,836 biogenic toxins.
• Advanced Biotransformation Prediction Model: Integrates over 8,000 biotransformation reaction rules derived from approximately 300,000 biochemical reactions curated from around 580,000 literature reports by more than 100 contributors over the past decade.
• Global Chemical Space Construction: Constructs a global chemical space encompassing roughly 550,000 toxins and putative metabolites, of which 94.7% are not previously reported.
• Case Study Application: Applied to citrinin metabolism in Trichoderma where it identified a novel metabolite.

Scientific Applications:

• Metabolome Discovery: Enables exploration of the "dark matter" of toxin metabolomes by predicting unreported metabolites.
• Enzyme Discovery: Facilitates identification of putative detoxification enzymes through linkage of predicted metabolites to experimental validation.
• Environmental and Safety Studies: Supports research on biogenic toxin dynamics in environmental contexts, food safety, and toxicology by providing predicted metabolite profiles.

Methodology:

Integrates literature-curated biochemical reactions and biotransformation reaction rules to generate a predicted metabolite space and link computational predictions to experimental validation.