To create the 23 mammal EPO alignment, the methodology had to be extended to include low-coverage genomes. The assemblies of low-coverage genomes are too fragmented, creating too many breakpoints in the Enredo graph, to use Enredo directly. The Enredo graph was therefore built using high-coverage genomes only. Low-coverage genomes were then mapped on the colinear regions using pairwise alignments to the human genome. For each low-coverage genome, the segments defined by the pairwise alignments were linked with stretches of N's to facilitate the process of building the final multiple sequence alignment. After the alignment has been obtained, the stretches of N's were removed.
