matchms

To identify unknown compounds from mass spectrometry spectra, you match them against reference libraries using spectral similarity scoring. This involves loading LC-MS/MS data, applying preprocessing to normalize peaks, and ranking candidates by tolerance-aware cosine similarity.

Spectral similarity scoring in metabolomics library search is the process of ranking unknown MS/MS spectra by comparing them to reference spectra. It uses tolerance-aware matching algorithms like cosine and modified cosine to calculate similarity scores for compound identification.

You preprocess MS/MS data for spectral matching by harmonizing metadata, normalizing peak intensities, filtering peaks, and enforcing quality thresholds. Standardization ensures consistent format handling when importing files like MGF, mzML, or MSP for analysis.

Yes, you can import mzML and MGF files for metabolomics compound identification. The process supports loading spectra from common formats including MGF, mzML, MSP, JSON, and pickle, allowing you to match unknowns against reference spectral libraries.

The best way to cluster metabolomics spectra after library searching is to apply standardization pipelines that harmonize metadata and normalize peaks. This allows you to compute precursor-aware similarity scores for spectral clustering across large datasets.

Spectral matching for compound identification requires precursor m/z tolerance to accurately compute precursor-aware similarity scores. Configurable peak filtering and tolerance-aware matching algorithms ensure accurate cosine-based ranking against reference libraries.