Table 1.

Performance of classifiers (AUC) on the human test set.

Sample	Classification method
	ENLR	SVM	RF
1	0.90	0.91	0.91
2	0.82	0.86	0.89
3	0.82	0.88	0.87
4	0.85	0.91	0.76
5	0.77	0.80	0.71
Average	0.83	0.87	0.83

Abbreviations: AUC, area under ROC curve, ROC, receiver operating characteristic; ENLR, Elastic Net-regularized Logistic Regression; RF, Random Forests; SVM, Support Vector Machine.

All 3 classifiers have an average AUC above 0.8. SVM achieved the best performance in the identification of condition-specific m⁶A sites.

The bold values in the table highlight the best performance achieved in each sample.