Mind the gap: upgrading genomes with Pacific Biosciences RS long-read sequencing technology.

English AC, Richards S, Han Y, Wang M, Vee V, Qu J, Qin X, Muzny DM, Reid JG, Worley KC, Gibbs RA. · 2012

Read Mind the gap: upgrading genomes with Pacific Biosciences RS long-read sequencing technology. by English AC, Richards S, Han Y, Wang M, Vee V, Qu J, Qin X, Muzny DM, Reid JG, Worley KC, Gibbs RA. (2012) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc3504050.txt.

Introduction

Draft genomes are full of holes, and Adam C. English and twelve colleagues present PBJelly in 2012, a tool that patches them with long reads from the Pacific Biosciences RS sequencer. Short-read assemblies stumble on repeats, biases, and variation, and manual finishing with older Sanger methods is slow and costly.

PBJelly guides long reads onto a draft assembly, filling gaps automatically. It also produces coordinate tables that carry existing annotations from the old draft onto the upgraded sequence, protecting years of prior work. The software is public, and the authors show the approach on real genomes. Long reads, they argue, turn finishing from a craft project into a routine automated step.

Catalog record

CREATOREnglish AC, Richards S, Han Y, Wang M, Vee V, Qu J, Qin X, Muzny DM, Reid JG, Worley KC, Gibbs RA.
YEAR2012
COLLECTIONPAPERS
FORMATTXT
FILE SIZE36 KB
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC3504050

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Mind the Gap: Upgrading Genomes with Pacific Biosciences RS Long-Read Sequencing Technology By Adam C. English, Stephen Richards, Yi Han, Min Wang, Vanesa Vee, Jiaxin Qu, Xiang Qin, Donna M. Muzny, Jeffrey G. Reid, Kim C. Worley, Richard A. Gibbs, Zhanjiang Liu Published in: PLoS ONE DOI: 10.1371/journal.pone.0047768 ABSTRACT Many genomes have been sequenced to high-quality draft status using Sanger capillary electrophoresis and/or newer short-read sequence data and whole genome assembly techniques. However, even the best draft genomes contain gaps and other imperfections due to limitations in the input data and the techniques used to build draft assemblies. Sequencing biases, repetitive genomic features, genomic polymorphism, and other complicating factors all come together to make some regions difficult or impossible to assemble. Traditionally, draft genomes were upgraded to “phase 3 finished” status using time-consuming and expensive Sanger-based manual finishing processes. For more facile assembly and automated finishing of draft genomes, we present here an automated approach to finishing using long-reads from the Pacific Biosciences RS (PacBio) platform. Our algorithm and associated software tool, PBJelly, (publicly available at https://sourceforge.net/projects/pb-jelly/) automates the finishing process using long sequence reads in a reference-guided assembly process. PBJelly also provides “lift-over” co-ordinate tables to easily port existing annotations to the upgraded assembly. Using PBJelly and long PacBio reads, we upgraded the draft genome sequences of a simulated Drosophila melanogaster, the version 2 draft Drosophila pseudoobscura, an assembly of the Assemblathon 2.0 budgerigar dataset, and a preliminary assembly of the Sooty mangabey. With 24× mapped coverage of PacBio long-reads, we addressed 99% of gaps and were able to close 69% and improve 12% of all gaps in D. pseudoobscura. With 4× mapped coverage of PacBio long-reads we saw reads address 63% of gaps in our budgerigar assembly, of which 32% were closed and 63% improved. With 6.8× mapped coverage of mangabey PacBio long-reads we addressed 97% of gaps and closed 66% of addressed gaps and improved 19%. The accuracy of gap closure was validated by comparison to Sanger sequencing on gaps from the original D. pseudoobscura draft assembly and shown to be dependent on initial reference quality. INTRODUCTION Genome finishing has become a lost art due to the expense of oligonucleotide directed Sanger sequencing relative to the low cost-per-base of second generation sequencing technologies. The first generation of large eukaryotic model organism genome sequencing projects, such as Drosophila melanogaster [1], Caenorhabditis elegans [2], Arabidopsis thaliana [3], human [4] , and mouse [5], all relied on a mapped bacterial artificial chromosome (BAC) approach. In the BAC approach, individual mapped BACs were shotgun sequenced, assembled, and manually finished before being pieced together creating the final, finished reference genome. Because of the prohibitive cost and labor required for BAC library creation, arraying, mapping, and preparation of subclone libraries from tens of thousands of BACs, these techniques fell out of favor. They were replaced by significantly less expensive and time-consuming whole genome assembly methods. Initially, assembly methods used relatively long (500–800 bp) shotgun Sanger reads with Overlap-Layout-Consensus assemblers [6]–[11]. Due to financial considerations, Sanger whole genome assemblies often used as few reads as possible, saving millions of dollars, but producing lower quality genomes using as little as 6× genome coverage, falling significantly short of the 10–15× required for high quality draft assemblies. Second-generation genome assemblies have been based on shorter read, massively parallel sequencing technologies and De Bruijn graph assembly techniques [12]–[15] Depending on the dataset quality, polymorphism, and repetitiveness of the target genome, both approaches generate draft genomes with contig N50 sizes ranging from 5 kb–200 kb. Assemblies with short contig N50 statistics suffer from having many gene models (the foundation of most biological research) with gaps, missing exons, genes split between scaffolds, or missing entirely. As an anecdotal example, consider the rhesus macaque – an important biomedical model organism with a large international research community. The current draft of the rhesus macaque genome contains sequence gaps in up to 20% of its gene models. Most other eukaryotic genomes larger than yeast are currently assembled only to draft genome quality and have similar problems of varying degree (Table 1). The scale of the unfinished genome problem will be compounded by new initiatives to sequence 10,000 vertebrate genomes (http://www.genome10k.org/), 5,000 arthropod genomes (http://arthropodgenomes.org/wiki/i5K) and 1,000 additional plant and animal genomes (http://ldl.genomics.org.cn/page/pa-research.jsp.) [Table: Gap numbers and size distributions for representative high quality draft assemblies of highly studied species.] Historically, several approaches have been used to upgrade draft genome sequences in a cost effective and automated manner. Early on, cosmids and BACs were assembled from forward only reads, and subclones pointing into gaps were then selected for reverse sequencing to reduce primer design costs and enable easier automation [16]. As capillary electrophoresis techniques evolved, Sanger reads as long as 1 kb were used to aid gap closure in an approach similar to ours (LI-COR Biosciences Lincoln NE). With the advent of massively-parallel short read technologies, paired reads of multiple insert sizes have been used to “reach” into gaps from unique contig sequences – for example Atlas-GapFill and the SOAP-denovo gap filler both use this approach. It is important to note the distinctions between genome upgrading & finishing, and simple re-assembly. While better assembly algorithms may allow incremental improvements in re-assembling a given data set, the resulting contigs and scaffolds will have no relationship to previous assembly versions, thus losing existing annotations. In contrast, genome upgrading fills gaps and upgrades low quality regions, preserving most of the assembled sequence and annotations. To fully exploit the improved assembly, annotations must be re-evaluated in regions with improved sequence and newly closed gaps, but this is unnecessary in the majority of the assembly. The Pacific Biosciences RS (PacBio) is the first sequencing technology to offer very long read lengths (average read lengths are 2–3 kb and >7 kb reads are not uncommon) without GC-bias or systematic errors, though it suffers the shortcomings of modest throughput and low accuracy (∼15% error rate skewed toward insertions). Despite these issues, extremely long and unbiased reads are uniquely suited for upgrading genomes. We developed a software tool (PBJelly) that uses PacBio reads to close gaps and preserve annotations. We applied PBJelly to four draft genome assemblies (one from simulated data), creating improved versions of these genomes and “lift-over” tables for annotation preservation. RESULTS Design and Implementation of PBJelly PBJelly is an automated pipeline for gap filling and genome improvement that aligns long sequence reads to draft assembles in order to close or improve captured gaps. PBJelly is currently applicable to PacBio RS reads, but it can be generalized to apply to any long-reads. PBJelly was designed and implemented with the goals of making genome upgrades automatic, fast, accurate, and reproducible. PBJelly trusts the input draft genome to be accurate and functions to improve what is already known about the genome, modifying the existing draft as little as possible by focusing on gaps and regions with missing and/or low-quality data. Furthermore, PBJelly verbosely logs all improvements to the draft genome, which enables identification and rejection of questionable gap fills, and production of an annotation co-ordinates lift-over table. Finally, PBJelly can be run on any cluster (we use Moab/Torque/PBS) parallelizing the gap filling process for rapid turn around. Figure 1a diagrams the PBJelly workflow.

Read the complete text