Integrated, Accurate and Multi-Environment Structural Variation Discovery from Whole Genome Sequencing Data with NGSEP

Structural Variants (SVs) are genomic differences between two or more samples that encompass large regions of their DNA. Whole genome sequencing (WGS) data can be effectively used to discover SVs, combining different algorithms specialized on different types of variants. However, integration of software tools for SV discovery is currently difficult because different tools are implemented in different languages, have different sets of requirements, and differ greatly in their parameter settings. We previously released NGSEP as a software tool that tightly integrates state-of-the-art algorithms for simultaneous discovery of SNVs, indels, and copy number variation (CNVs). Continuing this effort, we recently implemented in NGSEP three new algorithms for SV discovery: CNV-seq, for read-depth (RD) comparison between two samples; RDXplorer for detection of small CNVs; and a novel implementation of the read-pair (RP) and split-read (SR) strategies for detection of medium to large indels, including breakpoint resolution. Comparisons using benchmark datasets of yeast and rice shows that the combination of RD, RP and SR strategies integrated in NGSEP achieves superior accuracy and efficiency compared to individual approaches implemented in other currently available software packages. To facilitate the use of NGSEP we improved its documentation for command line usage and its integration with the Galaxy environment. We also made NGSEP available within the iPlant collaborative platform and the DNAnexus cloud platform for SV discovery in the 3000 rice genomes project. We expect that our continuous development efforts facilitate the use of NGSEP for a growing number of researchers in plant and animal genomics.