DNA sequence assembly and multiple sequence alignment by an Eulerian path approach.

Abstract

We describe an Eulerian path approach to the DNA fragment assembly that was originated by Idury and Waterman 1995, and then advanced by Pevzner et al. 2001b. This combinatorial approach bypasses the traditional “overlap-layout-consensus” approach and successfully resolved some of the troublesome repeats in practical assembly projects. The assembly results by the Eulerian path approach are accurate, and its computation is significantly more efficient than other assembly programs. As an extension, we use the Eulerian path idea to address the multiple sequence alignment problem. In particular, we have as a goal aligning thousands of sequences simultaneously, which is computationally exorbitant for all existing alignment algorithms. As a beginning, we focus on DNA sequence alignment. Our method can align hundreds of DNA sequences within minutes with high accuracy, and its computational time is linear to the number of sequences. We demonstrate its performance by alignments of simulated sequences and by an application in a resequencing project of Arabidopsis thaliana. Although having some weaknesses including aligning gap-rich regions, the Eulerian path approach is distinguished from other existing algorithms in solving either fragment assembly or multiple alignment