Figure 3

Exogean outperforms all other automatic methods confronted during EGASP. Histogram of the performances in terms of specificity (Sp, grey), sensitivity (Sn, white), and average between specificity and sensitivity ((Sn+Sn)/2, black) for all methods designed to predict protein-coding sequences in the EGASP competition. The values were provided by the EGASP organizers, and are based on predictions matching at least one coding sequence (CDS) exactly for each Havana annotated gene (known or novel). Themethods are ranked left to right from best to worse average between Sn and Sp.