Fig. 4: DeepCDM-assisted environmental non-targeted analysis. | Nature Communications

Fig. 4: DeepCDM-assisted environmental non-targeted analysis.

From: Deep learning prediction of electrospray ionization tandem mass spectra of chemically derived molecules

Fig. 4: DeepCDM-assisted environmental non-targeted analysis.

a Dansylation-specialized in silico spectral library (DnsBank) enabled confidence level 3 annotation of chemicals in liquid chromatography-mass spectrometry (LC-MS) based non-targeted analysis of leather industrial wastewater after dansylation. b Total annotation numbers of 3 wastewater samples obtained from DnsBank spectrum matching, SIRIUS 4 and Compound DiscovererTM 3.3 (CD 3.3) database searching. c Composition of tentative annotations in the untreated wastewater (UW), activate carbon adsorption (ACD) wastewater and ACD/ozonation wastewater obtained from DnsBank. d The uniform manifold approximation and projection (UMAP) of chemicals detected in the 3 wastewater samples. e Identity confirmation of 3-fluoro-5-hydroxybenzonitrile and ent-paroxol by matching tandem mass spectra with the dansylated chemical standards. f Concentration−response curves of Hep G2 cells incubating with 3-fluoro-5-hydroxybenzonitrile and ent-paroxol. Each concentration had 6 biological replicates. Data are presented as mean values ± standard deviation. Source data are provided as a Source Data file.

Back to article page