matchms — Mass Spectrometry Processing

What it does

Adds mass-spectrometry data processing capabilities: import/export mzML, MGF, MSP, JSON; harmonize metadata; apply filters and QC; compute spectral similarity (cosine, modified cosine, fingerprint-based); and build reproducible processing pipelines. Useful for metabolomics, library matching, and large-scale spectral comparisons.

When to use it

Use this skill when you need to preprocess mass spectrometry datasets, clean and harmonize metadata, filter noisy peaks, match unknown spectra against reference libraries, or compute spectral similarity scores for annotation and clustering. It fits workflows that ingest mzML/MGF/MSP files, run standardized filters, and output processed libraries for downstream analysis.

What's included

• Scripts: none detected in repo (has_references=true).
• References: comprehensive markdown references (filtering, similarity, importing/exporting, workflows) included in the skill folder.
• Instructions: step-by-step examples show how to load spectra, apply default_filters and intensity normalization, build a SpectrumProcessor pipeline, and calculate scores using CosineGreedy and ModifiedCosine. The SKILL.md includes code snippets for importing/exporting, filtering, pipeline construction, and spectrum visualization.

Compatible agents

Primarily useful for code-capable agents that can run or author Python: Copilot/Codex/GPT-code-style agents, Claude Code, and other agents able to produce/run Python data-science workflows.