Fig. 2: KGMN improves peak annotation accuracy. | Nature Communications

Fig. 2: KGMN improves peak annotation accuracy.

From: Metabolite annotation from knowns to unknowns through knowledge-guided multi-layer metabolic networking

Fig. 2

a Peak correlation subnetworks for metabolic features of M285T555 and M153T555. The M153T555 was correctly recognized as the in-source fragment of M285T555 by KGMN. The feature identifier is defined as the (nominal) mass and RT in seconds (e.g., M153T555: nominal mass 153, retention time 555 s). b The validation of M153T555 (m/z 153.0398) as an in-source fragment of M285T555 (m/z 285.0810) using the chemical standard of xanthosine. c–e The improved accuracy of known metabolite annotation using MetDNA1 and KGMN: c, statistics of annotation numbers; d, percentages of annotation coverage and correct/error rates; e correct and error rates among top n annotations. The x-axis represents the top n metabolite candidates for each peak. Source data are provided as a Source Data file.

Back to article page