Metabolite annotation and identification

NPClassifier

Deep neural-network tool that classifies natural-product structures into biosynthetic pathways, superclasses and structural classes.

Platform-independentWeb applicationAPIOpen source

Overview

NPClassifier assigns natural-product pathway, superclass and class labels from a supplied chemical structure. Within MS metabolomics it can organize already annotated compounds or candidate structures. It does not infer a structure directly from a mass spectrum.

What can this tool do?

Compound-class prediction

Common uses

  • Group candidate or annotated metabolites by natural-product chemical class.
  • Add structure-derived class labels to downstream metabolomics summaries.

Scientific fit

Primary function

Metabolite annotation and identification

Data compatibility

Inputs

Molecular structures

Outputs

Compound list or annotations

Input formats

SMILES or InChI, SMILES strings (InChI input is not asserted)

Output formats

JSON

Scope and considerations

Designed for

  • Biosynthetically informed classification of supplied natural-product structures.

Not designed for

  • Annotating an unknown MS/MS spectrum without a candidate structure.
  • Proving that a predicted class establishes biological origin or a biosynthetic route.

Important requirements

  • A valid molecular structure supplied as SMILES.
  • For local hosting, the model files and deployment dependencies described in the repository.

Considerations

  • Classification depends on the supplied structure; an incorrect structural annotation carries through to its class labels.
  • The natural-product ontology has its own class definitions and coverage; preserve that context when comparing classifications.

Access and documentation

The hosted classifier is publicly accessible. The repository specifies MIT for software and CC0 for its data, models and ontology. Local deployment is documented; the repository also states that queried structures are logged by the hosted service.

License: MIT

Evidence and review

Editorial review of official documentation, repository and the direct tool publication on 2026-09-10. No analysis job, independent benchmark or developer verification was performed.

Core publications

Resource papers, direct updates and tool-specific methods. Selection policy.

  1. Kim HW, Wang M, Leber CA, et al. NPClassifier: A Deep Neural Network-Based Structural Classification Tool for Natural Products. Journal of Natural Products 84, 2795–2807 (2021). doi:10.1021/acs.jnatprod.1c00399 PMID 34662515

Related tools

Entry history

Changes to this catalogue entry, not the tool's software release history. Acceptance and editorial review are separate.

  1. 2026-09-10 · First committed entry

    Add three GNPS2 Tools and expose complete related-tool lists

    View change · Entry at this revision

Full GitHub history