matchms

To identify metabolites from mass spectrometry data, import spectra from formats like MGF, MSP, mzML, or JSON, apply default filters to normalize metadata, and compute cosine similarity scores against a reference library to retrieve top matches.

Spectral matching compares query mass spectra against reference libraries using similarity scoring algorithms like CosineGreedy, ModifiedCosine, and NeutralLossesCosine to robustly annotate chemical structures and identify metabolites.

Yes, you can import and harmonize spectra from multiple formats including mzML, MGF, MSP, and JSON within the same reproducible pipeline to perform comprehensive library searching across diverse data sources.

The best way to normalize mass spectra metadata is to apply programmable filtering pipelines that harmonize metadata fields across imported formats, ensuring consistent annotation linking to chemical structures like InChI and InChIKey before comparison.

For library searching, use multi-metric scoring with CosineGreedy, ModifiedCosine, and NeutralLossesCosine to evaluate different structural alignment properties, providing robust metabolite identification across diverse spectral data.

Deriving fingerprints during metabolite annotation links matched spectra to chemical structures for integrated analyses, enabling structural comparison and validation alongside the similarity scores obtained from your library search.