A novel plant classification method based on similarities in chemical structures of metabolite contents obtained from the KNApSAcK database
Metabolite content refers to all small molecules that are the products or intermediates of metabolism within an organism. Similarities between metabolite contents may be used to investigate the evolution of and bioactivity-based relationships between plants. The KNApSAcK core database (DB) is an extensive plant-metabolite relationship database that is utilized in multifaceted plant research, such as the identification of metabolites and other bioinformatic and systems biology research. We herein present a novel method to classify plants based on structural similarities in their metabolite contents obtained from the KNApSAcK Core DB (http://kanaya.naist.jp/KNApSAcK/). The genetic make-up of an organism results in its physical characteristics. The genetic information of an organism is used to control omics levels such as the formation of proteins and the regulation of metabolism. Therefore, we hypothesize that the metabolite content of an organism is a kind of overall signature, and that this information may be utilized to classify a group of organisms. We initially selected a set of plants such that each of them produced at least several dozen metabolites. The Tanimoto coefficients (TCs) of the structural similarities between all possible pairs of metabolites in our data were then investigated, and comprise the background population of TCs. TCs were also assessed for all possible unique pairs of metabolites corresponding to each pair of plants, which was translated into a plant-plant similarity score using thresholds based on the background population of TCs. Plant-plant similarity scores were then used to classify plants. The classifications obtained were compared with the NCBI taxonomy. The resulting classifications were then explained in the context of the phylogeny and bioactivity of plants.
Liu, K., Altaf-Ul-Amin, Md., Abdullah, A.A., Morita, A.H., Shiraishi, M. and Kanaya, S. 2017. A novel plant classification method based on similarities in chemical structures of metabolite contents obtained from the KNApSAcK database. Acta Hort. (ISHS) 1169:139-150
species-metabolite relationships, Tanimoto coefficients, plant taxonomy