Figure 4

Divergence Distance Estimation. Divergences estimated from tool alignments are overestimated at short divergence distances and underestimated at large divergence distances while divergences estimated from true simulated alignments are accurate to large divergence distances. A: Mean divergence distance estimated from simulated alignments and tool alignments for four species trees was measured as a function of total true divergence distance. B: Mean divergence estimation error (Estimate – True/True) for four species trees was measured as a function of total true divergence distance. C: Divergence estimation error from tool alignments is not correlated with alignment error. Mean divergence estimation error for four species trees was measured as a function of mean alignment error. D: Divergence estimation error varies across branches in a tree and is best for leaf-to-leaf alignments and worst for node-to-node alignments. Mean divergence estimation error along branches of equal true divergence from two, three and four species Mlagan alignments was measured as a function of true divergence distance.