Table 9. Ablation analysis of FSA on simulated mammalian genomic DNA.

FSA options	Blanchette et al.
	(Acc/Sn/PPV)
(default)	0.53/0.32/0.93
–exonerate	0.83/0.77/0.94
–exonerate –minscore 50	0.83/0.78/0.94
–exonerate –refinement 0	0.82/0.76/0.93
–exonerate –noindel2	0.78/0.72/0.94

Ablation analysis of FSA on the simulated mammalian DNA of Table 3 : Comparisons of the accuracies (Acc), sensitivities (Sn) and positive predictive values (PPV) of FSA with different components enabled or disabled. We tested the effectiveness of components related to anchor annealing for aligning long sequences, including using anchors from MUMmer and exonerate and changing the minimum acceptable score for an exonerate anchor (the default is –minscore 100). These results clearly show that while using only MUMs for anchoring (the default mode) gives a high positive predictive value, inexact matches must be used to obtain high sensitivity on very long or distant nonfunctional sequences lacking the local constraints which give rise to MUMs across species in functional (e.g., coding) sequence.