A comprehensive review of computational methods for predicting DNA N⁴-methylcytosine sites

N4-methylcytosine (4mC) is a distinct form of DNA methylation that plays a critical role in various biological processes by protecting bacterial DNA from degradation and participating in the regulation of gene expression. With advances in technology, computational approaches have increasingly replaced traditional experimental methods, which are often associated with high costs, prolonged processing times, and labor-intensive workflows. Over the past five years, a growing number of machine learning (ML) and deep learning (DL) models have been developed to predict 4mC sites. In this review, we provide a systematic overview of these computational methods, focusing on model architectures and comparing the strengths and limitations of ML- and DL-based approaches. To facilitate future tool development, we have collected and organized commonly used databases and benchmark datasets relevant to 4mC prediction. In addition, we compared several recently proposed methods to highlight their respective strengths and capabilities. Finally, we highlight the current challenges and opportunities in the field, aiming to facilitate the development of more accurate and robust predictive frameworks for 4mC methylation.