Overview
MSBERT learns MS/MS embeddings using masked and contrastive learning. Examples compute spectrum similarities and integrate the scoring method into matchms.
What can this tool do?
Common uses
- Embed tandem spectra for similarity scoring and reference-library matching.
Scientific fit
Primary function
Spectral processing and comparison
Data compatibility
Inputs
Mass spectra, Spectral library
Outputs
Human-readable report
Input formats
MSP tandem spectra and trained MSBERT weights
Output formats
Embedding vectors and pairwise spectrum similarity scores
Scope and considerations
Designed for
- MSBERT learns MS/MS embeddings using masked and contrastive learning. Examples compute spectrum similarities and integrate the scoring method into matchms.
Not designed for
- De novo structure generation or evidence of identity from similarity alone.
Important requirements
- Python/PyTorch environment and pretrained model; reference spectra for library search.
Considerations
- Training and evaluation use curated GNPS data; assess performance for the instrument and chemistry of your study.
- A documented Streamlit interface is available; hosted service operation was not tested.
Access and documentation
MIT license verified in the official source repository. Training datasets and model downloads may have separate terms. No execution or independent performance benchmark.
License: MIT
Evidence and review
Enveda intake checked against official documentation, license information and primary DOI/title/year on 2026-10-22. Editorial review only; no execution or independent benchmark.
Core publications
Resource papers, direct updates and tool-specific methods. Selection policy.
- MSBERT: Embedding Tandem Mass Spectra into Chemically Rational Space by Mask Learning and Contrastive Learning. Analytical Chemistry (2024). doi:10.1021/acs.analchem.4c02426
Related tools
Entry history
Changes to this catalogue entry, not the tool's software release history. Acceptance and editorial review are separate.
- 2026-10-23 · First committed entry
Add ten reviewed Enveda tools with input and applicability limits
View change · Entry at this revision