An efficient approach to BAC based assembly of complex genomes

Abstract

BackgroundThere has been an exponential growth in the number of genome sequencing projects since the introduction of next generation DNA sequencing technologies. Genome projects have increasingly involved assembly of whole genome data which produces inferior assemblies compared to traditional Sanger sequencing of genomic fragments cloned into bacterial artificial chromosomes (BACs). While whole genome shotgun sequencing using next generation sequencing (NGS) is relatively fast and inexpensive, this method is extremely challenging for highly complex genomes, where polyploidy or high repeat content confounds accurate assembly, or where a highly accurate ‘gold’ reference is required. Several attempts have been made to improve genome sequencing approaches by incorporating NGS methods, to variable success.ResultsWe present the application of a novel BAC sequencing approach which combines indexed pools of BACs, Illumina paired read sequencing, a sequence assembler specifically designed for complex BAC assembly, and a custom bioinformatics pipeline. We demonstrate this method by sequencing and assembling BAC cloned fragments from bread wheat and sugarcane genomes.ConclusionsWe demonstrate that our assembly approach is accurate, robust, cost effective and scalable, with applications for complete genome sequencing in large and complex genomes.Electronic supplementary materialThe online version of this article (doi:10.1186/s13007-016-0107-9) contains supplementary material, which is available to authorized users.

“Targeted Sequencing by Gene Synteny,” a New Strategy for Polyploid Species: Sequencing and Physical Structure of a Complex Sugarcane Region

“…The total length of all the assembled contigs was 8.94 Mb, with an N50 contig length of 91.5 kb and a GC content of 44.74%. The N50 value was higher than that obtained during wheat BAC sequencing using only long reads generated by PacBio, which exhibited a mean N50 of 80 kb (Visendi et al, 2016 ).…”

“…Several studies using bacterial artificial chromosomes (BACs), involving either individual BAC assembly (de Setta et al, 2014 ; Vilela et al, 2017 ) or pooled strategies (Okura et al, 2016 ; Visendi et al, 2016 ), have been reported. In both cases, the applied sequencing strategies are based on the selection of non-overlapping BAC clones.…”

A single polyploidization event at the origin of the tetraploid genome of Coffea arabica is responsible for the extremely low genetic variation in wild and cultivated germplasm

“…To tackle this, we could have opted to isolate chromosomes through flow cytometry 29 but we could not consider this technology due to lack of appropriate genetic stocks that would allow for easy chromosome isolation. We thus opted for a hierarchical sequencing approach of BAC pooling where the likelihood of two homoeologous fragments occurring in the same pool was extremely low 30,31 , leading to dramatic reduction of the genome complexity in the assemblies of each pool. In this paper, we show a successful application of this genome assembly for studying Arabica diversity.…”

Integrated physical map of bread wheat chromosome arm 7DS to facilitate gene cloning and comparative studies

“…A total of 4608 MTP BAC clones were pair-end sequenced using a pooling strategy in which 96 pools, each consisting of four non-overlapping clones, were indexed and sequenced on a single lane of the Illumina HiSeq2000 platform [32]. DNA sequences were de-multiplexed and assembled using SASSY as described in detail by Visendi et al [36]. DNA sequence deconvolution was supported by BAC-end sequences (BES), obtained by Sanger sequencing of all MTP clones from both ends, and by utilizing overlaps between BAC clones in physical map contigs.…”

“…DNA sequence deconvolution was supported by BAC-end sequences (BES), obtained by Sanger sequencing of all MTP clones from both ends, and by utilizing overlaps between BAC clones in physical map contigs. Mate-pair data was obtained by Illumina sequencing MTPplate pools (384 clones per pool), and mate-pair reads were then applied to build scaffolds by SSPACE as described in Visendi et al [36]. Alignments of sequence scaffolds to the 7DS BNG map were performed in IrysView 2.1.1.…”