ASRmiRNA: Abiotic Stress-Responsive miRNA Prediction in Plants by Using Machine Learning Algorithms with Pseudo K-Tuple Nucleotide Compositional Features

Abstract

MicroRNAs (miRNAs) play a significant role in plant response to different abiotic stresses. Thus, identification of abiotic stress-responsive miRNAs holds immense importance in crop breeding programmes to develop cultivars resistant to abiotic stresses. In this study, we developed a machine learning-based computational method for prediction of miRNAs associated with abiotic stresses. Three types of datasets were used for prediction, i.e., miRNA, Pre-miRNA, and Pre-miRNA + miRNA. The pseudo K-tuple nucleotide compositional features were generated for each sequence to transform the sequence data into numeric feature vectors. Support vector machine (SVM) was employed for prediction. The area under receiver operating characteristics curve (auROC) of 70.21, 69.71, 77.94 and area under precision-recall curve (auPRC) of 69.96, 65.64, 77.32 percentages were obtained for miRNA, Pre-miRNA, and Pre-miRNA + miRNA datasets, respectively. Overall prediction accuracies for the independent test set were 62.33, 64.85, 69.21 percentages, respectively, for the three datasets. The SVM also achieved higher accuracy than other learning methods such as random forest, extreme gradient boosting, and adaptive boosting. To implement our method with ease, an online prediction server “ASRmiRNA” has been developed. The proposed approach is believed to supplement the existing effort for identification of abiotic stress-responsive miRNAs and Pre-miRNAs.

Discovering Common miRNA Signatures Underlying Female-Specific Cancers via a Machine Learning Approach Driven by the Cancer Hallmark ERBB

“…On these selected features we applied a Support Vector Machine (SVM) algorithm to execute a Tumor vs. Normal binary classification. We selected this machine learning method, based on its wide usage in the field of transcriptomic and miRNA classification works, as reported in the literature [ 43 , 44 , 49 , 50 , 51 ]. Other methods (i.e., Logistic regression, Boosted Logistic Regression, Regression with LASSO penalty, Elastic Net, Random Forest, Neural Networks using Model Averaging (avNNet, and the “nnet” package)) produced similar results in AUC and other metrics, compared with SVM (the AUC obtained from the tested ML methods ranging from 0.911 to 0,97 compared to an AUC of 0.931 by SVM) as reported in Table S3 .…”

“…Hence, we performed an integrated bioinformatics analysis by combining in silico miRNAs discovered via a machine learning approach from TCGA-BRCA, -OV, and -UCEC datasets and the miRTarBase knowledge for in silico identification of miRNAs targeting ERBB isoforms. In the field of transcriptomic and miRNA classification, many studies adopted the SVM machine learning method because it outperformed the others [ 43 , 44 , 49 , 50 , 51 ]. We also chose an SVM approach, since other methods (i.e., logistic regression, Boosted Logistic Regression, Quantile Regression with LASSO penalty, Elastic Net, Random Forest, avNNet, and the “nnet” package) produced relatively similar results for AUC and other metrics compared with SVM.…”

Machine Learning-Based Identification of Mating Type and Metalaxyl Response in Phytophthora infestans Using SSR Markers

“…Although the classification accuracy can be deemed adequate in the machine learning context, particularly for the random forest model, it is debatable whether the accuracy is sufficient to guide decision-making. Moreover, in the plants within some machine learning studies, such as for the prediction of miRNAs associated with abiotic stresses [82], an accuracy of less than 70% was reported for some models. In humans, an accuracy of 65% to 70% was reported for diagnostic prediction of autism [83].…”

A review of artificial intelligence-assisted omics techniques in plant defense: current trends and future directions

“…Similarly, Meher et al. (2022a) contributed to the field by developing a ML-based method for predicting miRNAs responsive to abiotic stresses.…”

“… Meher et al. (2022a) employed computational methods and machine learning (ML) to streamline the identification of abiotic stress-responsive genes (SRGs) across various stress conditions, achieving accuracy levels of 60% to 78% with the SVM model.…”