Metabolite annotation and identification

SEISMiQ

Generate candidate molecular structures from tandem spectra and a supplied molecular formula.

LC-MSCommand lineOpen source

Overview

SEISMiQ is a generative structure-elucidation model developed for drug-substance impurities. Its released prediction workflow and CASMI checkpoints also support small-molecule MS/MS candidate generation.

What can this tool do?

Structure elucidation

Common uses

  • Generate candidate molecular structures from tandem spectra and a supplied molecular formula.

Scientific fit

Primary function

Metabolite annotation and identification

Data compatibility

Inputs

Mass spectra, Compound list or annotations

Outputs

Molecular structures

Input formats

JSON challenges containing an MS/MS peak list and sum formula

Output formats

CSV candidate SMILES and model scores

Scope and considerations

Designed for

  • SEISMiQ is a generative structure-elucidation model developed for drug-substance impurities. Its released prediction workflow and CASMI checkpoints also support small-molecule MS/MS candidate generation.

Not designed for

  • Unconstrained identification from spectra alone or proof of metabolite identity.

Important requirements

  • Python/Poetry environment, compatible model checkpoint and tokenizer.
  • MS/MS spectrum and molecular formula; sufficient memory or a suitable GPU.

Considerations

  • Training and fine-tuning domains affect applicability to biological metabolites; do not infer universal metabolomics performance.
  • Optional substructure prompts and adduct assumptions must be appropriate for the sample.

Access and documentation

License verified in the official repository. Code licensing does not establish the terms for every separately downloaded model or dataset.

License: Apache-2.0

Evidence and review

Enveda intake candidate checked against official documentation, license information and primary publication on 2026-10-21. Editorial review only; no execution or independent benchmark.

Core publications

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

  1. Enhancing automated drug substance impurity structure elucidation from tandem mass spectra through transfer learning and domain knowledge. Digital Discovery (2025). doi:10.1039/d5dd00115c

    Final peer-reviewed article replaces the differently titled intake preprint.

Related tools

Entry history

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

  1. 2026-10-22 · First committed entry

    Add MAMSI and SEISMiQ with explicit applicability limits

    View change · Entry at this revision

Full GitHub history