Combined network analysis and machine learning allows the prediction of metabolic pathways from tomato metabolomics data

• Published in: Communications biology 2019-06, Vol.2 (1), p.214-214, Article 214
• Main Authors: Toubiana, David, Puzis, Rami, Wen, Lingling, Sikron, Noga, Kurmanbayeva, Assylay, Soltabayeva, Aigerim, del Mar Rubio Wilhelmi, Maria, Sade, Nir, Fait, Aaron, Sagi, Moshe, Blumwald, Eduardo, Elovici, Yuval
• Format: Article
• Language: English
• Subjects: 631/114/1305; 631/114/2397; 631/114/2408; 631/449/447; 631/61/320; Alanine; Biology; Biomedical and Life Sciences; Crop improvement; Drug development; Indoles; Learning algorithms; Life Sciences; Machine Learning; Melibiose; Metabolic Networks and Pathways; Metabolic pathways; Metabolism; Metabolites; Metabolomics; Metabolomics - methods; Pericarp; Pyruvic acid; Solanum lycopersicum - metabolism; Tryptophan
• ISSN: 2399-3642; 2399-3642
• DOI: 10.1038/s42003-019-0440-4

The identification and understanding of metabolic pathways is a key aspect in crop improvement and drug design. The common approach for their detection is based on gene annotation and ontology. Correlation-based network analysis, where metabolites are arranged into network formation, is used as a complentary tool. Here, we demonstrate the detection of metabolic pathways based on correlation-based network analysis combined with machine-learning techniques. Metabolites of known tomato pathways, non-tomato pathways, and random sets of metabolites were mapped as subgraphs onto metabolite correlation networks of the tomato pericarp. Network features were computed for each subgraph, generating a machine-learning model. The model predicted the presence of the β-alanine-degradation-I, tryptophan-degradation-VII-via-indole-3-pyruvate (yet unknown to plants), the β-alanine-biosynthesis-III, and the melibiose-degradation pathway, although melibiose was not part of the networks. In vivo assays validated the presence of the melibiose-degradation pathway. For the remaining pathways only some of the genes encoding regulatory enzymes were detected. David Toubiana et al. report the combination of correlation-based network analysis and machine-learning techniques to detect metabolic pathways from quantitative metabolic data. They demonstrate the utility of their method by studying metabolic pathways in tomato.