Fig. 4.

Performance comparison among iREAD, IRFinder and their combined score on the simulated 15 M and 30 M RNA-seq datasets with known ground-truth intron retention events in terms of AUC. The ROC curves of both methods based on the GS1(FPKM≥0.1), GS2 (FPKM≥0.3) and GS3(FPKM≥0.5) gold standard for the 15 M dataset is shown in (a), (b) and (c), respectively. Similarly, the ROC curves of both methods based on the GS1, GS2 and GS3 gold standard for the 30 M dataset are shown in (d), (e) and (f), respectively