omicverse.metabol.fetch_chebi_compounds

omicverse.metabol.fetch_chebi_compounds#

omicverse.metabol.fetch_chebi_compounds(*, cache=True, refresh=False)[source]#

Build a compound master table from ChEBI’s flat-file TSVs.

Downloads + joins three ChEBI distributions from the public EBI FTP (over HTTPS):

  • compounds.tsv.gz — ChEBI ID → canonical name

  • chemical_data.tsv.gz — monoisotopic mass + formula

  • database_accession.tsv.gz — HMDB / KEGG / LipidMaps xrefs

Total download is ~15 MB; the joined parquet cache persists at ~/.cache/omicverse/metabol/chebi_compounds.parquet. This is the substrate annotate_peaks() uses for mummichog mass matching.

Returns:

Columns: chebi_id, name, formula, mw (monoisotopic, float), kegg, hmdb, lipidmaps. Rows without a monoisotopic mass are dropped.

Return type:

pd.DataFrame

Contents