Machine learning improves our knowledge about miRNA functions towards plant abiotic stresses.

During the last two decades, human has increased his knowledge about the role of miRNAs and their target genes in plant stress response. Biotic and abiotic stresses result in simultaneous tissue-specific up/down-regulation of several miRNAs. In this study, for the first time, feature selection algorithms have been used to investigate the contribution of individual plant miRNAs in Arabidopsis thaliana response towards different levels of several abiotic stresses including drought, salinity, cold, and heat. Results of information theory-based feature selection revealed that miRNA-169, miRNA-159, miRNA-396, and miRNA-393 had the highest contributions to plant response towards drought, salinity, cold, and heat, respectively. Furthermore, regression models, i.e., decision tree (DT), support vector machines (SVMs), and Naïve Bayes (NB) were used to predict the plant stress by having the plant miRNAs' concentration. SVM with Gaussian kernel was capable of predicting plant stress (R = 0.96) considering miRNA concentrations as input features. Findings of this study prove the performance of machine learning as a promising tool to investigate some aspects of miRNAs' contribution to plant stress responses that have been undiscovered until today.