Fig. 4. AllCCS improves known metabolite annotation.

a Candidate reduction with addition of CCS match demonstrated using validation set 2; b percentages of rank improvement for correct candidates with addition of CCS match to multi-dimensional match; three in-silico MS/MS prediction tools (CFM-ID, MetFrag, and MS-FINDER) were used to generate predicted MS/MS spectra; c an example of rank improvement for the compound 6-Hydroxycoumarin with addition of CCS match in multi-dimensional match; d schematic illustration of multi-dimensional annotation of metabolites in biological samples; e candidate reduction with addition of CCS match in metabolite annotation of MEF cell samples; f the candidate reductions in different biological samples; the CCS match tolerance was set as 4%; g percentages for candidate reduction with multi-dimensional match for each feature in MEF cell samples; h the hypoxanthine was successfully annotated with multi-dimensional properties; MS-FINDER was used to generate predicted MS/MS spectra for metabolite candidates. Source data are provided as a Source Data file.