A Computational Study of Potential miRNA-Disease Association Inference Based on Ensemble Learning and Kernel Ridge...

Area of Science:

• Bioinformatics
• Genomics
• Computational Biology

Background:

• MicroRNAs (miRNAs) play crucial roles in biological processes and human diseases.
• Experimental identification of miRNA-disease associations is resource-intensive.
• Computational methods are needed to complement experimental approaches.

Purpose of the Study:

• To develop a novel computational model for predicting miRNA-disease associations.
• To integrate diverse similarity measures for enhanced prediction accuracy.
• To validate the model's performance through rigorous cross-validation and case studies.

Main Methods:

• Developed Ensemble of Kernel Ridge Regression based MiRNA-Disease Association prediction (EKRRMDA) model.
• Integrated disease semantic similarity, miRNA functional similarity, and Gaussian interaction profile kernel similarity.
• Employed ensemble learning and feature dimensionality reduction with Kernel Ridge Regression base classifiers.

Main Results:

• EKRRMDA achieved high performance in cross-validation, with AUCs of 0.9314 (global LOOCV) and 0.8618 (local LOOCV).
• Achieved an average AUC of 0.9275 ± 0.0008 in 5-fold cross-validation.
• Case studies showed high validation rates for top predicted miRNAs associated with Esophageal Neoplasms (90%), Kidney Neoplasms (86%), Lymphoma (86%), Lung Neoplasms (98%), and Breast Neoplasms (96%).

Conclusions:

• EKRRMDA is an effective computational tool for predicting miRNA-disease associations.
• The model's high accuracy and validation rates support its utility in biological research and clinical applications.
• This approach can accelerate the discovery of novel miRNA-disease links, contributing to disease prevention and treatment strategies.