UM-PPS

UM-PPS predicts microbial catabolic pathways and biotransformations of organic compounds using biotransformation rules derived from the University of Minnesota Biocatalysis/Biodegradation Database (UMBBD) and the scientific literature.

Key Features:

• Biotransformation rule-based prediction: Applies a comprehensive set of biotransformation rules to predict successive microbial transformations of organic compounds.
• Rule provenance: Prediction rules are derived from the University of Minnesota Biocatalysis/Biodegradation Database (UMBBD) and corroborated by published scientific literature.
• Functional group identification: Identifies functional groups within organic molecules that are likely targets for microbial degradation.
• Aerobic Likelihood Assessment: Uses expert-assigned aerobic likelihood categories (Very Likely, Likely, Neutral, Unlikely, Very Unlikely) to prioritize transformations under aerobic conditions.
• Filtering by likelihood: Supports filtering of predicted transformations based on aerobic likelihood to focus on more probable pathways.
• Relative Reasoning: Allows certain rules to inhibit others to reduce unlikely or conflicting predictions.
• Super rules: Aggregates individual reaction rules into super rules that represent multiple contiguous reactions forming coherent sub-pathways.
• Rule set refinement and stringency control: Continuously modifies and increases rule stringency to reduce extraneous predicted pathways and improve prediction quality.

Scientific Applications:

• Environmental biotechnology: Prediction of microbial degradation pathways for assessment and engineering of biodegradation processes.
• Toxicology: Anticipation of biotransformation products relevant to toxicity and exposure assessment.
• Bioremediation: Identification of plausible microbial catabolic routes for remediation planning of contaminated sites.
• Synthetic biology: Inference of enzymatic transformations and sub-pathways useful for pathway design and metabolic engineering.
• Microbial catabolism research: Generation of hypothesis-driven predictions to guide experimental investigation of microbial degradation mechanisms.

Methodology:

Identifies functional groups in input molecules, applies biotransformation rules from UMBBD and literature, assigns expert aerobic likelihood categories, employs relative reasoning to allow rule inhibition, converts applicable rules into super rules representing contiguous reactions, and iteratively refines the rule set to increase stringency.