Figure 9

Performance computed in absence of similar proteins at different similarity levels. Figures (A) and (B) show the precision recall curves computed after excluding from the sets of similar structures homologs (without regarding the species) sharing ≥ 95% of sequence identity with the query proteins. Similarly, figures (C) and (D) show the performances after excluding proteins sharing ≥ 50% sequence identity, and (E) and (F) display the results after filtering out proteins with sequence identity ≥ 30%. The precision-recall curves corresponding to the DS56bound dataset are shown at (A), (C), and (E), and the results computed using the DS56Unbound dataset are labeled as (B), (D), and (F). Figures (E) and (F) were computed using 55 and 52 proteins respectively given that PredUs could not find structural elements for the protein chain 1ynt-L.