m5C-TNKmer: Identification of 5-Methylated Base Cytosine of Ribonucleic Acid Using Supervised Machine Learning Techniques

Abstract

5-Methylcytosine (m⁵C) is a widely recognized epigenetic modification in ribonucleic acid (RNA), catalyzed by methyltransferases. This modification is crucial for various biological functions. While the role of m⁵C in deoxyribonucleic acid (DNA) has been extensively studied, its role in RNA is still in its early stages of exploration. Accurate and systematic detection and classification of m⁵C sites in RNA remain challenging tasks. Machine learning techniques offer an efficient alternative to traditional laboratory methods for identifying m⁵C sites in Homo sapiens. This study introduces a novel computational model m⁵C-TNKmer, which utilizes k-mer feature extraction to enhance the identification of m⁵C sites in RNA sequences. Four sub-datasets derived from the primary dataset Di-nucleotide (DNC), Tri-nucleotide (TNC), Tetra-nucleotide (Tetra-NC), and Penta-nucleotide (Penta-NC) were used to train the model. The results demonstrated that m⁵C-TNKmer achieved an impressive accuracy of 96.15%. This model provides a powerful tool for scientists to accurately identify RNA m⁵C sites, contributing to a deeper understanding of genetic functions and regulatory mechanisms.