Figure 7
From: Efficient pairwise RNA structure prediction and alignment using sequence alignment constraints
Performance of constrained Sankoff implementations on test sets binned by percent identity. Each bin represents 10% identity and contains 10 randomly selected pairwise comparisons. The top panel (A) shows the alignment accuracy of Consan (circles, black line), Dynalign (squares, blue line), Stemloc (triangles, red line), FOLDALIGN (stars, orange line), and PMcomp (diamonds, green line). The lower panel (B) shows the structure prediction accuarcy of the same implementations.