I used a sliding window of 50 bp to measure variation in 25-bp steps along the length of the aligned coding sequence (note that because gaps are excluded, the length of the region compared varies among alignments).
In [ 2], the Monge-Elkan score was reported to perform the best among alignment-based measures in most cases, if the score matrix {5, 3, -3} is used, and if the score is scaled to the interval [0, 1].
This result is contrary to results reported in [ 2], which states: the Monge-Elkan score is the best among alignment-based measures, the Levenshtein distance is considerably worse than the Monge-Elkan score, and SoftTFIDF is the best overall distance measure for their dataset.
Therefore ties among alignments with the same edge correctness are broken in favor of those with the highest overall bit score.
The size of regions with the same topology was comparable among alignments simulated with ILS and those simulated with HGT overall.
We noticed that alignments with higher percent identity (Additional file 1: Figure S1 8) or lower e-values (Additional file 1: Figure S9 16) were more frequently observed among alignments to parental KO-families than non-parental KO-families.
The requirement for a two base difference among alignments was used so that a SNP allele in one DG sequence would not be confused with a DG sequence that maps to a different site.
To test this hypothesis we calculated maximum likelihood estimates of rates of non-synonymous (dN) and synonymous substitution (dS) among alignments of Rhox2, 3 and 4 paralogues using a codon-based model of sequence evolution.
The IUPAC ambiguity code recruited eleven additional letters from the Roman alphabet to indicate positions in a consensus sequence where two, three, or four bases were found among alignments of structurally and functionally homologous sequences.
The variance among alignments in GC content is generally lower in mouse than in pig and human, but mostly so for coding sequences, followed by the UTR and intron regions (Table 2).
