We then used MuMMer 2.3[ 31]to align the contig sets for each strain, using an all-against-all alignment to identify contigs that were similarly constructed between the assemblers.
When larger datasets become available, we expect the total running time of an all-against-all alignment to increase more than quadratic with dataset size: We have shown above that a large fraction of the total running time stems from a few 'hard' alignments which, in turn, correspond to a few trees in the dataset that are large and, in particular, have high out degrees.
For each query, all alignments to distinct chromosomes with alignment length ≥100 base pairs (bp) and identity ≥95% were collected and processed.
All alignments to that nearest neighbor, including lower scoring ones, were considered as portions of the complete alignment.
We confirm these findings even when we limit from all alignments to the best alignments only (just as above) (Additional file 1: Figures S8 S10).
Therefore, a redundancy collapse required all alignments to be global with respect to one of the two sequences and semiglobal with respect to the other.
A random sequence of equal length and composition was included in all alignments to show pairwise percent identities that are not significantly different from random identity.
We will additionally upload all alignments to protein domain databases such as Pfam, so that researchers can access them more easily.
In our PR analysis, a PR curve is constructed for each method on each alignment, and all the PR curves for a method are averaged across all alignments to obtain its overall curve.
Because breseq uses read mapping options that report all alignments to sequence repeats in the reference genome, pairs of alignments for the same read that match different reference locations may yield equivalent junction sequences but different junction descriptions.
Sequence coverage, gauged by comparing the total length of all alignments to the total length of all sequences in the OM set, was maximized by choosing a k-mer size of 39 (see Additional file 8: Figure S1).
