Adaboost-SVM-based probability algorithm for the prediction of all mature miRNA sites based on structured-sequence...

Area of Science:

• Bioinformatics
• Molecular Biology
• Computational Biology

Background:

• MicroRNAs (miRNAs) play crucial roles in biological processes and diseases.
• Current computational methods for mature miRNA identification face challenges in feature extraction and class imbalance.

Purpose of the Study:

• To develop an accurate computational classifier, miRFinder, for mature miRNA identification.
• To address limitations of existing methods in feature extraction and class imbalance.

Main Methods:

• Utilized structured-sequence features for precise miRNA biological feature extraction.
• Employed K-means clustering for center of mass near distance training to mitigate class imbalance.
• Constructed the classifier using the AdaBoost-SVM algorithm, focusing on incorrectly classified samples.

Main Results:

• The miRFinder classifier demonstrated high efficacy in identifying mature miRNAs compared to other methods.
• Structured-sequence features and AdaBoost-SVM significantly improved classifier performance.
• The approach effectively handled class imbalance issues inherent in miRNA identification.

Conclusions:

• miRFinder offers an accurate and effective alternative for mature miRNA identification.
• The proposed methods enhance the reliability of computational miRNA analysis.
• This tool can advance research in miRNA-related biological processes and diseases.