In addition, the identity percentage cutoff for a significant alignment was set to be 100 for the last cycle and 95 for all other cycles of Megablast [ 32].
Our goal here is to find the valid disjoint fragment chain C d (of length B≤ K) over Fset with the highest possible 'score' with respect to the scoring function fscore and transition penalty function fpenalty given as Here, the 'transcript gap penalty function' (according to the constant gap penalty scheme described in our nucleotide-level alignment formulation) is set to be P t (dist)= Cgap*dist.
The alignment parameters were set to be as similar as possible across the different tools when permitting.
If alignment blocks are set to be short with longer intervals, the blocks would be frequently allocated to a single different genealogy, leading to the precise estimation of speciation times and population sizes.
After 20 replications of this procedure, we found that if alignment blocks are set to be short with longer intervals, speciation times and population sizes were estimated precisely (fig. 1 C – H).
Bowtie2 options are used where alignment scores are set to be equal to the number of base matches minus three times the number of base mismatches (down-weighted if they are lower quality base calls), with a gap open penalty of 2 and a gap extension penalty of 3. A staged alignment procedure is employed to accelerate read mapping and alignment.
For every alignment, 20% of sites were set to be invariant.
Sequencer, alignment, and variant calling parameters were set to be identical in the analysis of all exome-based sequencing platforms (WES and ACE) with the exception of the target capture region used, which is specific to each platform.
The MAC threshold is set to 0.5 and the P-value threshold for suboptimal alignment is set to 0.1.
If no alignment is found, the distance is set to 1.
By default BLAT is set to allow such alignment extensions, but MegaBLAST is not, resulting in the type of differences between alignments presented in Figure 3B.
