Figure 3.

The performances of different m⁶A site predictors on the gold standard dataset and the benchmarking datasets. (A) Prediction results on the golden standard dataset. The gene identifiers and site positions are in lines with the original publication by Liu et al. (42). Experimental reference sites and predicted sites are denoted in the E and P columns, respectively. Experimentally verified m⁶A sites and non-m⁶A sites are indicated by deep red and grey boxes, respectively. Predicted very high confidence m⁶A sites, high confidence m⁶A sites and non-m⁶A sites are indicated by red, purple and grey boxes, respectively. (B) The ROC curve illustrating the performances on the mammalian benchmarking dataset. (C) The precision-recall curve illustrating the performances on the mammalian benchmarking dataset. (D) The ROC curve illustrating the performances on the yeast benchmarking dataset. (E) The precision-recall curve illustrating the performances on the yeast benchmarking dataset.