Figure 4.

Ranking diagrams of various aligners. The 12 aligners for cDNA and CDS (a) or the seven aligners for protein (b) are ranked according to exon-level accuracies averaged over the three species for each of the six identity levels of simulated datasets or seven pairs of cross-species datasets. Succeeding aligners are connected by a shaded box if the difference in their performance is not significant (P ≥ 0.01) or by a shaded triangle if the difference is weakly significant (0.001 ≤ P < 0.01).