Whole Genome Alignments Based On Local Alignments

Given the genomes (DNA) of two related species, one of the whole genome alignment tasks is to locate regions on the genomes that possibly contain genes conserved over the two species. Most of the existing whole genome alignment software tools, MUMmer (Delcher et al., 2002; Kurtz et al., 2004), MaxMincluster (Wong et al., 2004), and MSS (Chan et al., 2004), generate Maximal Unique Matches (MUMs) directly from the DNA sequences as the input of the whole genome alignments. Yet the sensitivity of the previous methods is quite low. Motivated by existing heuristic-based software tools to find the homology regions between two species (PatternHunter), we initiate the study of using the results of local alignments to produce MUMs in order to improve the preciseness of the input MUMs. In this project we propose four strategies to generate MUMs from local alignments. And we find that only two of the four strategies can produce considerable results. It is exciting to see that when compared with MUMs generated directly from sequences on real virus data sets, even the brute-force strategy highly improves the preciseness of the output after processing MSS, which is better than MUMmer and MaxMincluster. At the same time, the new strategy uncovers slightly more conserved genes.