Fig. 4
From: VARUS: sampling complementary RNA reads from the sequence read archive
Relative Annotation Accuracy. The right side shows the difference in whole-genome annotation accuracy in percent. The F1-measure of coding exon accuracy (the harmonic mean of sensitivity and specificity) was used, whereby either annotation was compared to the respective reference annotation (see Supplementary Table 2). The left shows the corresponding input data set sizes. The average and mean number of spots chosen by the manual method are 100 and 77 million, respectively, and larger than the 50 million spots downloaded by VARUS. Averaging over the species, the F1 accuracy of BRAKER is 0.62% higher when RNA-Seq is selected by VARUS rather than manual