“…These predictors typically make use of machine learning algorithms to learn from available data to perform novel predictions and gain new insights. Recently, a variety of machine learning algorithms are useful for this goal, such as support vector machine (SVM) ( Chen et al, 2017 ; He et al, 2019 ; Wei et al, 2019a , b ; Lv et al, 2020b ; Zhao et al, 2020 ), random forest (RF) ( Hasan et al, 2020a , b ; Lv et al, 2020a ; Alghamdi et al, 2021 ; Zulfiqar et al, 2021a ), Markov model (MM) ( Yang et al, 2020 ), and the combined or ensemble methods ( Gong and Fan, 2019 ; Manavalan et al, 2019a , b ; Tang et al, 2020 ; Li et al, 2021 ), extreme gradient boosting (XGBoost) ( Wang et al, 2021 ) and Laplacian Regularized Sparse Representation ( Ding et al, 2021 ). As shown in Supplementary Table 1 , SVM is the most widely used traditional machine learning algorithms in the model development and method comparison for 4mC prediction, followed by RF.…”