Figure 1.

Overview of the Benchmark approach. The main contributions of this study are summarized in this figure. First, we provide human labeled benchmark ORF sets for four organisms. To the best of our knowledge, these datasets were not used for the development of ORF prediction tools so far, and therefore provide a valuable resource for the community (left blue box). Second, we provide a complete workflow where future novel tools can be easily tested (blue box, right). Finally, we compare the predictive performance and secondary measures of eight tools with our benchmark set of prokaryotic ORFs and corresponding Ribo-seq data.