Fig. 5
From: On the evaluation of the fidelity of supervised classifiers in the prediction of chimeric RNAs
The fidelity of chRNA classifiers using simulated RNA-seq data. The radar chart of using distinct chimera scores (0.7, 0.9, 0.95, 0.99, and 1) on CRAC (CracScore) and its impact on prediction performance. CracScore to 0.7 is a good compromising between precision (PPR) 100bp (a), PPR 200bp (b), and false discovery rate (FDR) (c). Moreover, the higher the CracScore is the higher is the precision though with the cost of increasing the FDR at the read level