Tools · Cheminformatics & Structure

Metabolite Enumerator

Predict likely phase-I and phase-II metabolites of your molecules, ranked, with the biotransformation pathway behind each one.

Updated October 1, 2026

On this page

Prices, workflows, and method papers Open in the app

The Metabolite Enumerator proposes what the body might turn a molecule into. It applies SyGMa's curated human biotransformation rules: phase I (functionalization, such as hydroxylation or dealkylation) and phase II (conjugation, such as glucuronidation or sulfation). Each candidate keeps its parent, the rules that produced it, and a score, and the candidates are ordinary molecules you can pass to ADMET Predictor, structural alerts, or docking to look for metabolite-specific liabilities.

These are hypotheses, not observed metabolites. It runs on CPU.

How it works

SyGMa applies phase-I rules for phase1_cycles generations (0 to 2, default 1), then phase-II rules for phase2_cycles (0 or 1, default 1). A candidate's score is the product of the empirical probabilities of the rules along its pathway: it ranks candidates for one parent, but is not a calibrated probability. Each structure appears once per parent, with its best-scoring pathway; fragments much smaller than the parent are dropped.

A second phase-I generation produces many more candidates (often ten times as many) and takes correspondingly longer.

Inputs

Up to 1,000 parent molecules as SMILES, SDF or MOL blocks, or InChI. Rules match the structure as given, so standardize tautomers first with Molecule Standardizer if they matter. Options: max_metabolites_per_parent (1 to 200, default 25, the top-scoring ones) and min_score (default 0).

Outputs

File	Contents
`metabolites.smi`, `metabolites.sdf`	Every kept metabolite with its id
`pathways.csv`	One row per metabolite: parent, SMILES, InChIKey, score, rank, generation, phases (`I`, `II`, or `I+II`), and the rules applied in order

A parent that fails to parse, or has no metabolite above min_score, gets a status row instead; the job carries on. When the cap cuts a parent's list, its rows are marked truncated. The same metabolite reached from two parents is listed under each, with duplicate_of pointing to the first.

Screen the metabolites with ADMET Predictor or Drug-likeness & Structural Alerts.

Run it from the API

Submit with Submit a job and the job_type below. Price it first with Estimate job reservation cost: submitting reserves that amount from your wallet, and the charge settles at the actual runtime.