Fast and Efficient Rmap Assembly using the Bi-labelled de Bruijn Graph

Abstract

Genome wide optical maps are high resolution restriction maps that give a unique numeric representation to a genome. They are produced by assembling hundreds of thousands of single molecule optical maps, which are called Rmaps. Unfortunately, there exists very few choices for assembling Rmap data. There exists only one publicly-available non-proprietary method for assembly and one proprietary method that is available via an executable. Furthermore, the publicly-available method, by Valouev et al. (2006), follows the overlap-layout-consensus (OLC) paradigm, and therefore, is unable to scale for relatively large genomes. The algorithm behind the proprietary method, Bionano Genomics' Solve, is largely unknown. In this paper, we extend the definition of bi-labels in the paired de Bruijn graph to the context of optical mapping data, and present the first de Bruijn graph based method for Rmap assembly. We implement our approach, which we refer to as Rmapper, and compare its performance against the assembler of Valouev et al. (2006) and Solve by Bionano Genomics on data from three genomes - E. coli, human, and climbing perch fish (Anabas Testudineus). Our method was able to successfully run on all three genomes. The method of Valouev et al.(2006) only successfully ran on E. coli. Moreover, on the human genome Rmapper was at least 130 times faster than Bionano Solve, used five times less memory and produced the highest genome fraction with zero mis-assemblies. Our software, RMAPPER is written in C++ and is publicly available under GNU General Public License at https://github.com/kingufl/Rmapper.

Finding Overlapping Rmaps via Gaussian Mixture Model Clustering

“…Since pairwise alignment of Rmaps is a critical step for Rmap assembly, the lack of an efficient solution to this problem has inhibited the development of Rmap assemblers. Currently there exists only two non-proprietary methods for Rmap assembly: the assembly method of Valouev et al [26], and a de Bruijn graph-based assembler called rmapper [18]. The former uses the alignment method of Valouev et al to first compute all pairwise alignments and uses a threshold alignment score to identify reliable overlaps among Rmaps.…”

“…Dynamic programming remains the only robust method for finding pairwise alignments between Rmaps as existing alignment methods largely focus on finding alignments between assembled genome wide optical maps [13,15]. Unfortunately, dynamic programming is computationally expensive and is unable to scale to even moderately large sized genomes, such as the human genome [18]. Therefore, all existing error correction methods (cOMet and Elmeri) use heuristics to filter out pairs of Rmaps that are likely to not have significant pairwise alignment, and then find alignments between the remaining pairs [19,22].…”