Overview
MSGO generates candidate molecular structures from tandem spectra with a model trained using simulated spectrum-structure pairs. The published implementation provides PFAS and lipid model routes with beam-search candidate outputs.
What can this tool do?
Common uses
- A research implementation for de novo molecular-structure generation from mass-spectral evidence using pseudodata-trained machine learning.
Scientific fit
Primary function
Metabolite annotation and identification
Other functions
Machine learning and prediction
Data compatibility
Inputs
Mass spectra
Outputs
Molecular structures
Input formats
Example-structured tandem-spectrum CSVs and trained model checkpoints
Output formats
CSV of ranked generated molecular-structure candidates
Scope and considerations
Designed for
- A research implementation for de novo molecular-structure generation from mass-spectral evidence using pseudodata-trained machine learning.
Not designed for
- A universal structure generator with validated coverage of all chemical classes or confirmed identifications from top-ranked SMILES alone.
Important requirements
- The documented Python/PyTorch environment, matching model weights and correctly formatted spectrum CSVs; use the polarity and model appropriate to the chemical domain.
Considerations
- Pseudodata and restricted PFAS/lipid training domains limit transfer. Some model/data assets use external download services; inspect accessibility and precursor/formula consistency before interpreting generated candidates.
Access and documentation
MIT source distribution. External datasets, dependent packages and any hosted routes retain their own terms; software operation was not tested.
License: MIT
Evidence and review
Enveda intake rechecked against official repository documentation, package/license metadata and primary publication identity on 2026-10-27. Editorial source review only; no software execution or independent benchmark. See docs/enveda-expansion-review-2026-10-27-batch2.md.
Core publications
Resource papers, direct updates and tool-specific methods. Selection policy.
- Pseudodata-based molecular structure generator to reveal unknown chemicals. Nature Machine Intelligence (2025). doi:10.1038/s42256-025-01140-5
Related tools
Entry history
Changes to this catalogue entry, not the tool's software release history. Acceptance and editorial review are separate.
- 2026-10-27 · First committed entry
Add 24 reviewed Enveda tools in a separate second batch
View change · Entry at this revision