Long-read Assembly Research Articles

Recompleting the Caenorhabditis elegans genome

Caenorhabditis elegans was the first multicellular eukaryotic genome sequenced to apparent completion. Although this assembly employed a standard C. elegans strain (N2), it used sequence data from several laboratories, with DNA propagated in bacteria and yeast. Thus, the N2 assembly has many differences from any C. elegans available today. To provide a more accurate C. elegans genome, we performed long-read assembly of VC2010, a modern strain derived from N2. Our VC2010 assembly has 99.98% identity to N2 but with an additional 1.8 Mb including tandem repeat expansions and genome duplications. For 116 structural discrepancies between N2 and VC2010, 97 structures matching VC2010 (84%) were also found in two outgroup strains, implying deficiencies in N2. Over 98% of N2 genes encoded unchanged products in VC2010; moreover, we predicted ≥53 new genes in VC2010. The recompleted genome of C. elegans should be a valuable resource for genetics, genomics, and systems biology.

Resolving repeat families with long reads

BackgroundDraft quality genomes for a multitude of organisms have become common due to the advancement of genome assemblers using long-read technologies with high error rates. Although current assemblies are substantially more contiguous than assemblies based on short reads, complete chromosomal assemblies are still challenging. Interspersed repeat families with multiple copy versions dominate the contig and scaffold ends of current long-read assemblies for complex genomes. These repeat families generally remain unresolved, as existing algorithmic solutions either do not scale to large copy numbers or can not handle the current high read error rates.ResultsWe propose novel repeat resolution methods for large interspersed repeat families and assess their accuracy on simulated data sets with various distinct repeat structures and on drosophila melanogaster transposons. Additionally, we compare our methods to an existing long read repeat resolution tool and show the improved accuracy of our method.ConclusionsOur results demonstrate the applicability of our methods for the improvement of the contiguity of genome assemblies.

Complete Assembly of Escherichia coli Sequence Type 131 Genomes Using Long Reads Demonstrates Antibiotic Resistance Gene Variation within Diverse Plasmid and Chromosomal Contexts.

The incidence of infections caused by extraintestinal Escherichia coli (ExPEC) is rising globally, which is a major public health concern. ExPEC strains that are resistant to antimicrobials have been associated with excess mortality, prolonged hospital stays, and higher health care costs. E. coli sequence type 131 (ST131) is a major ExPEC clonal group worldwide, with variable plasmid composition, and has an array of genes enabling antimicrobial resistance (AMR). ST131 isolates frequently encode the AMR genes blaCTX-M-14, blaCTX-M-15, and blaCTX-M-27, which are often rearranged, amplified, and translocated by mobile genetic elements (MGEs). Short DNA reads do not fully resolve the architecture of repetitive elements on plasmids to allow MGE structures encoding blaCTX-M genes to be fully determined. Here, we performed long-read sequencing to decipher the genome structures of six E. coli ST131 isolates from six patients. Most long-read assemblies generated entire chromosomes and plasmids as single contigs, in contrast to more fragmented assemblies created with short reads alone. The long-read assemblies highlighted diverse accessory genomes with blaCTX-M-15, blaCTX-M-14, and blaCTX-M-27 genes identified in three, one, and one isolates, respectively. One sample had no blaCTX-M gene. Two samples had chromosomal blaCTX-M-14 and blaCTX-M-15 genes, and the latter was at three distinct locations, likely transposed by the adjacent MGEs: ISEcp1, IS903B, and Tn2 This study showed that AMR genes exist in multiple different chromosomal and plasmid contexts, even between closely related isolates within a clonal group such as E. coli ST131.IMPORTANCE Drug-resistant bacteria are a major cause of illness worldwide, and a specific subtype called Escherichia coli ST131 causes a significant number of these infections. ST131 bacteria become resistant to treatments by modifying their DNA and by transferring genes among one another via large packages of genes called plasmids, like a game of pass-the-parcel. Tackling infections more effectively requires a better understanding of what plasmids are being exchanged and their exact contents. To achieve this, we applied new high-resolution DNA sequencing technology to six ST131 samples from infected patients and compared the output to that of an existing approach. A combination of methods shows that drug resistance genes on plasmids are highly mobile because they can jump into ST131's chromosomes. We found that the plasmids are very elastic and undergo extensive rearrangements even in closely related samples. This application of DNA sequencing technologies illustrates at a new level the highly dynamic nature of ST131 genomes.

TSD: A Computational Tool To Study the Complex Structural Variants Using PacBio Targeted Sequencing Data.

PacBio sequencing is a powerful approach to study DNA or RNA sequences in a longer scope. It is especially useful in exploring the complex structural variants generated by random integration or multiple rearrangement of endogenous or exogenous sequences. Here, we present a tool, TSD, for complex structural variant discovery using PacBio targeted sequencing data. It allows researchers to identify and visualize the genomic structures of targeted sequences by unlimited splitting, alignment and assembly of long PacBio reads. Application to the sequencing data derived from an HBV integrated human cell line(PLC/PRF/5) indicated that TSD could recover the full profile of HBV integration events, especially for the regions with the complex human-HBV genome integrations and multiple HBV rearrangements. Compared to other long read analysis tools, TSD showed a better performance for detecting complex genomic structural variants. TSD is publicly available at: https://github.com/menggf/tsd.

Whole-Genome Sequence of an Isogenic Haploid Strain, Saccharomyces cerevisiae IR-2idA30(MATa), Established from the Industrial Diploid Strain IR-2.

We present the draft genome sequence of an isogenic haploid strain, IR-2idA30(MAT a), established from Saccharomyces cerevisiae IR-2. Assembly of long reads and previously obtained contigs from the genome of diploid IR-2 resulted in 50 contigs, and the variations and sequencing errors were corrected by short reads.

Targeted, High-Resolution RNA Sequencing of Non-coding Genomic Regions Associated With Neuropsychiatric Functions.

The human brain is one of the last frontiers of biomedical research. Genome-wide association studies (GWAS) have succeeded in identifying thousands of haplotype blocks associated with a range of neuropsychiatric traits, including disorders such as schizophrenia, Alzheimer’s and Parkinson’s disease. However, the majority of single nucleotide polymorphisms (SNPs) that mark these haplotype blocks fall within non-coding regions of the genome, hindering their functional validation. While some of these GWAS loci may contain cis-acting regulatory DNA elements such as enhancers, we hypothesized that many are also transcribed into non-coding RNAs that are missing from publicly available transcriptome annotations. Here, we use targeted RNA capture (‘RNA CaptureSeq’) in combination with nanopore long-read cDNA sequencing to transcriptionally profile 1,023 haplotype blocks across the genome containing non-coding GWAS SNPs associated with neuropsychiatric traits, using post-mortem human brain tissue from three neurologically healthy donors. We find that the majority (62%) of targeted haplotype blocks, including 13% of intergenic blocks, are transcribed into novel, multi-exonic RNAs, most of which are not yet recorded in GENCODE annotations. We validated our findings with short-read RNA-seq, providing orthogonal confirmation of novel splice junctions and enabling a quantitative assessment of the long-read assemblies. Many novel transcripts are supported by independent evidence of transcription including cap analysis of gene expression (CAGE) data and epigenetic marks, and some show signs of potential functional roles. We present these transcriptomes as a preliminary atlas of non-coding transcription in human brain that can be used to connect neurological phenotypes with gene expression.

Assembly of long, error-prone reads using repeat graphs.

Accurate genome assembly is hampered by repetitive regions. Although long single molecule sequencing reads are better able to resolve genomic repeats than short-read data, most long-read assembly algorithms do not provide the repeat characterization necessary for producing optimal assemblies. Here, we present Flye, a long-read assembly algorithm that generates arbitrary paths in an unknown repeat graph, called disjointigs, and constructs an accurate repeat graph from these error-riddled disjointigs. We benchmark Flye against five state-of-the-art assemblers and show that it generates better or comparable assemblies, while being an order of magnitude faster. Flye nearly doubled the contiguity of the human genome assembly (as measured by the NGA50 assembly quality metric) compared with existing assemblers.

Graph analysis of fragmented long-read bacterial genome assemblies.

Long-read genome assembly tools are expected to reconstruct bacterial genomes nearly perfectly; however, they still produce fragmented assemblies in some cases. It would be beneficial to understand whether these cases are intrinsically impossible to resolve, or if assemblers are at fault, implying that genomes could be refined or even finished with little to no additional experimental cost. We propose a set of computational techniques to assist inspection of fragmented bacterial genome assemblies, through careful analysis of assembly graphs. By finding paths of overlapping raw reads between pairs of contigs, we recover potential short-range connections between contigs that were lost during the assembly process. We show that our procedure recovers 45% of missing contig adjacencies in fragmented Canu assemblies, on samples from the NCTC bacterial sequencing project. We also observe that a simple procedure based on enumerating weighted Hamiltonian cycles can suggest likely contig orderings. In our tests, the correct contig order is ranked first in half of the cases and within the top-three predictions in nearly all evaluated cases, providing a direction for finishing fragmented long-read assemblies. https://gitlab.inria.fr/pmarijon/knot . Supplementary data are available at Bioinformatics online.

Kermit: linkage map guided long read assembly

Background With long reads getting even longer and cheaper, large scale sequencing projects can be accomplished without short reads at an affordable cost. Due to the high error rates and less mature tools, de novo assembly of long reads is still challenging and often results in a large collection of contigs. Dense linkage maps are collections of markers whose location on the genome is approximately known. Therefore they provide long range information that has the potential to greatly aid in de novo assembly. Previously linkage maps have been used to detect misassemblies and to manually order contigs. However, no fully automated tools exist to incorporate linkage maps in assembly but instead large amounts of manual labour is needed to order the contigs into chromosomes.Results We formulate the genome assembly problem in the presence of linkage maps and present the first method for guided genome assembly using linkage maps. Our method is based on an additional cleaning step added to the assembly. We show that it can simplify the underlying assembly graph, resulting in more contiguous assemblies and reducing the amount of misassemblies when compared to de novo assembly.Conclusions We present the first method to integrate linkage maps directly into genome assembly. With a modest increase in runtime, our method improves contiguity and correctness of genome assembly.

Giant Transposons in Eukaryotes: Is Bigger Better?

Transposable elements (TEs) are ubiquitous in both prokaryotes and eukaryotes, and the dynamic character of their interaction with host genomes brings about numerous evolutionary innovations and shapes genome structure and function in a multitude of ways. In traditional classification systems, TEs are often being depicted in simplistic ways, based primarily on the key enzymes required for transposition, such as transposases/recombinases and reverse transcriptases. Recent progress in whole-genome sequencing and long-read assembly, combined with expansion of the familiar range of model organisms, resulted in identification of unprecedentedly long transposable units spanning dozens or even hundreds of kilobases, initially in prokaryotic and more recently in eukaryotic systems. Here, we focus on such oversized eukaryotic TEs, including retrotransposons and DNA transposons, outline their complex and often combinatorial nature and closely intertwined relationship with viruses, and discuss their potential for participating in transfer of long stretches of DNA in eukaryotes.

Hybrid assembly of ultra-long Nanopore reads augmented with 10x-Genomics contigs: Demonstrated with a human genome

The 3rd generation of sequencing (3GS) technologies generate ultra-long reads (up to 1 Mb), which makes it possible to eliminate gaps and effectively resolve repeats in genome assembly. However, the 3GS technologies suffer from the high base-level error rates (15%–40%) and high sequencing costs. To address these issues, the hybrid assembly strategy, which utilizes both 3GS reads and inexpensive NGS (next generation sequencing) short reads, was invented. Here, we use 10×-Genomics® technology, which integrates a novel bar-coding strategy with Illumina® NGS with an advantage of revealing long-range sequence information, to replace common NGS short reads for hybrid assembly of long erroneous 3GS reads. We demonstrate the feasibility of integrating the 3GS with 10×-Genomics technologies for a new strategy of hybrid de novo genome assembly by utilizing DBG2OLC and Sparc software packages, previously developed by the authors for regular hybrid assembly. Using a human genome as an example, we show that with only 7× coverage of ultra-long Nanopore® reads, augmented with 10× reads, our approach achieved nearly the same level of quality, compared with non-hybrid assembly with 35× coverage of Nanopore reads. Compared with the assembly with 10×-Genomics reads alone, our assembly is gapless with slightly high cost. These results suggest that our new hybrid assembly with ultra-long 3GS reads augmented with 10×-Genomics reads offers a low-cost (less than ¼ the cost of the non-hybrid assembly) and computationally light-weighted (only took 109 calendar hours with peak memory-usage = 61GB on a dual-CPU office workstation) solution for extending the wide applications of the 3GS technologies.

Long-read sequence and assembly of segmental duplications.

We developed a computational method based on polyploid phasing of long sequence reads to resolve collapsed regions of segmental duplications within genome assemblies. The approach, Segmental Duplication Assembler (SDA), constructs graphs where paralogous sequence variants define the nodes and long-read sequences provide attraction and repulsion edges allowing us to partition and assemble long reads corresponding to distinct paralogs. We apply it to single-molecule, real-time sequence data from three human genomes and recover 33–79 Mbp of duplications where approximately half of the loci are diverged (<99.8%) when compared to the reference genome. We show that the corresponding sequence is highly accurate (>99.9%) and that the diverged sequence corresponds to copy number variable paralogs that are absent from the human reference. Our method can be applied to other complex genomes to resolve the last gene-rich gaps, improve duplicate gene annotation, and better understand copy number variant genetic diversity at the base-pair level.

Resolving the complex Bordetella pertussis genome using barcoded nanopore sequencing.

The genome of Bordetella pertussis is complex, with high G+C content and many repeats, each longer than 1000 bp. Long-read sequencing offers the opportunity to produce single-contig B. pertussis assemblies using sequencing reads which are longer than the repetitive sections, with the potential to reveal genomic features which were previously unobservable in multi-contig assemblies produced by short-read sequencing alone. We used an R9.4 MinION flow cell and barcoding to sequence five B. pertussis strains in a single sequencing run. We then trialled combinations of the many nanopore user community-built long-read analysis tools to establish the current optimal assembly pipeline for B. pertussis genome sequences. This pipeline produced closed genome sequences for four strains, allowing visualization of inter-strain genomic rearrangement. Read mapping to the Tohama I reference genome suggests that the remaining strain contains an ultra-long duplicated region (almost 200 kbp), which was not resolved by our pipeline; further investigation also revealed that a second strain that was seemingly resolved by our pipeline may contain an even longer duplication, albeit in a small subset of cells. We have therefore demonstrated the ability to resolve the structure of several B. pertussis strains per single barcoded nanopore flow cell, but the genomes with highest complexity (e.g. very large duplicated regions) remain only partially resolved using the standard library preparation and will require an alternative library preparation method. For full strain characterization, we recommend hybrid assembly of long and short reads together; for comparison of genome arrangement, assembly using long reads alone is sufficient.

Fast Short Read De-Novo Assembly Using Overlap-Layout-Consensus Approach.

The de-novo genome assembly is a challenging computational problem for which several pipelines have been developed. The advent of long-read sequencing technology has resulted in a new set of algorithmic approaches for the assembly process. In this work, we identify that one of these new and fast long-read assembly techniques (using Minimap2 and Miniasm) can be modified for the short-read assembly process. This possibility motivated us to customize a long-read assembly approach for applications in a short-read assembly scenario. Here, we compare and contrast our proposed de-novo assembly pipeline (MiniSR) with three other recently developed programs for the assembly of bacterial and small eukaryotic genomes. We have documented two trade-offs: one between speed and accuracy and the other between contiguity and base-calling errors. Our proposed assembly pipeline shows a good balance in these trade-offs. The resulting pipeline is 6 and 2.2 times faster than the short-read assemblers Spades and SGA, respectively. MiniSR generates assemblies of superior N50 and NGA50 to SGA, although assemblies are less complete and accurate than those from Spades. A third tool, SOAPdenovo2, is as fast as our proposed pipeline but had poorer assembly quality.

Functional and Genome Sequence-Driven Characterization of tal Effector Gene Repertoires Reveals Novel Variants With Altered Specificities in Closely Related Malian Xanthomonas oryzae pv. oryzae Strains.

Rice bacterial leaf blight (BLB) is caused by Xanthomonas oryzae pv. oryzae (Xoo) which injects Transcription Activator-Like Effectors (TALEs) into the host cell to modulate the expression of target disease susceptibility genes. Xoo major-virulence TALEs universally target susceptibility genes of the SWEET sugar transporter family. TALE-unresponsive alleles of OsSWEET genes have been identified in the rice germplasm or created by genome editing and confer resistance to BLB. In recent years, BLB has become one of the major biotic constraints to rice cultivation in Mali. To inform the deployment of alternative sources of resistance in this country, rice lines carrying alleles of OsSWEET14 unresponsive to either TalF (formerly Tal5) or TalC, two important TALEs previously identified in West African Xoo, were challenged with a panel of strains recently isolated in Mali and were found to remain susceptible to these isolates. The characterization of TALE repertoires revealed that talF and talC specific molecular markers were simultaneously present in all surveyed Malian strains, suggesting that the corresponding TALEs are broadly deployed by Malian Xoo to redundantly target the OsSWEET14 gene promoter. Consistent with this, the capacity of most Malian Xoo to induce OsSWEET14 was unaffected by either talC- or talF-unresponsive alleles of this gene. Long-read sequencing and assembly of eight Malian Xoo genomes confirmed the widespread occurrence of active TalF and TalC variants and provided a detailed insight into the diversity of TALE repertoires. All sequenced strains shared nine evolutionary related tal effector genes. Notably, a new TalF variant that is unable to induce OsSWEET14 was identified. Furthermore, two distinct TalB variants were shown to have lost the ability to simultaneously induce two susceptibility genes as previously reported for the founding members of this group from strains MAI1 and BAI3. Yet, both new TalB variants retained the ability to induce one or the other of the two susceptibility genes. These results reveal molecular and functional differences in tal repertoires and will be important for the sustainable deployment of broad-spectrum and durable resistance to BLB in West Africa.

The unique evolution of the pig LRC, a single KIR but expansion of LILR and a novel Ig receptor family

The leukocyte receptor complex (LRC) encodes numerous immunoglobulin (Ig)-like receptors involved in innate immunity. These include the killer-cell Ig-like receptors (KIR) and the leukocyte Ig-like receptors (LILR) which can be polymorphic and vary greatly in number between species. Using the recent long-read genome assembly, Sscrofa11.1, we have characterized the porcine LRC on chromosome 6. We identified a ~ 197-kb region containing numerous LILR genes that were missing in previous assemblies. Out of 17 such LILR genes and fragments, six encode functional proteins, of which three are inhibitory and three are activating, while the majority of pseudogenes had the potential to encode activating receptors. Elsewhere in the LRC, between FCAR and GP6, we identified a novel gene that encodes two Ig-like domains and a long inhibitory intracellular tail. Comparison with two other porcine assemblies revealed a second, nearly identical, non-functional gene encoding a short intracellular tail with ambiguous function. These novel genes were found in a diverse range of mammalian species, including a pseudogene in humans, and typically consist of a single long-tailed receptor and a variable number of short-tailed receptors. Using porcine transcriptome data, both the novel inhibitory gene and the LILR were highly expressed in peripheral blood, while the single KIR gene, KIR2DL1, was either very poorly expressed or not at all. These observations are a prerequisite for improved understanding of immune cell functions in the pig and other species.

The complete replicons of 16 Ensifer meliloti strains offer insights into intra- and inter-replicon gene transfer, transposon-associated loci, and repeat elements.

Ensifer meliloti (formerly Rhizobium meliloti and Sinorhizobium meliloti) is a model bacterium for understanding legume–rhizobial symbioses. The tripartite genome of E. meliloti consists of a chromosome, pSymA and pSymB, and in some instances strain-specific accessory plasmids. The majority of previous sequencing studies have relied on the use of assemblies generated from short read sequencing, which leads to gaps and assembly errors. Here we used PacBio-based, long-read assemblies and were able to assemble, de novo, complete circular replicons. In this study, we sequenced, de novo-assembled and analysed 10 E. meliloti strains. Sequence comparisons were also done with data from six previously published genomes. We identified genome differences between the replicons, including mol% G+C and gene content, nucleotide repeats, and transposon-associated loci. Additionally, genomic rearrangements both within and between replicons were identified, providing insight into evolutionary processes at the structural level. There were few cases of inter-replicon gene transfer of core genes between the main replicons. Accessory plasmids were more similar to pSymA than to either pSymB or the chromosome, with respect to gene content, transposon content and G+C content. In our population, the accessory plasmids appeared to share an open genome with pSymA, which contains many nodulation- and nitrogen fixation-related genes. This may explain previous observations that horizontal gene transfer has a greater effect on the content of pSymA than pSymB, or the chromosome, and why some rhizobia show unstable nodulation phenotypes on legume hosts.

Genomes of 13 domesticated and wild rice relatives highlight genetic conservation, turnover and innovation across the genus Oryza

The genus Oryza is a model system for the study of molecular evolution over time scales ranging from a few thousand to 15 million years. Using 13 reference genomes spanning the Oryza species tree, we show that despite few large-scale chromosomal rearrangements rapid species diversification is mirrored by lineage-specific emergence and turnover of many novel elements, including transposons, and potential new coding and noncoding genes. Our study resolves controversial areas of the Oryza phylogeny, showing a complex history of introgression among different chromosomes in the young ‘AA’ subclade containing the two domesticated species. This study highlights the prevalence of functionally coupled disease resistance genes and identifies many new haplotypes of potential use for future crop protection. Finally, this study marks a milestone in modern rice research with the release of a complete long-read assembly of IR 8 ‘Miracle Rice’, which relieved famine and drove the Green Revolution in Asia 50 years ago.

BAUM: improving genome assembly by adaptive unique mapping and local overlap-layout-consensus approach.

It is highly desirable to assemble genomes of high continuity and consistency at low cost. The current bottleneck of draft genome continuity using the second generation sequencing (SGS) reads is primarily caused by uncertainty among repetitive sequences. Even though the single-molecule real-time sequencing technology is very promising to overcome the uncertainty issue, its relatively high cost and error rate add burden on budget or computation. Many long-read assemblers take the overlap-layout-consensus (OLC) paradigm, which is less sensitive to sequencing errors, heterozygosity and variability of coverage. However, current assemblers of SGS data do not sufficiently take advantage of the OLC approach. Aiming at minimizing uncertainty, the proposed method BAUM, breaks the whole genome into regions by adaptive unique mapping; then the local OLC is used to assemble each region in parallel. BAUM can (i) perform reference-assisted assembly based on the genome of a close species (ii) or improve the results of existing assemblies that are obtained based on short or long sequencing reads. The tests on two eukaryote genomes, a wild rice Oryza longistaminata and a parrot Melopsittacus undulatus, show that BAUM achieved substantial improvement on genome size and continuity. Besides, BAUM reconstructed a considerable amount of repetitive regions that failed to be assembled by existing short read assemblers. We also propose statistical approaches to control the uncertainty in different steps of BAUM. http://www.zhanyuwang.xin/wordpress/index.php/2017/07/21/baum. Supplementary data are available at Bioinformatics online.

MinION-based long-read sequencing and assembly extends the Caenorhabditis elegans reference genome.

Advances in long-read single molecule sequencing have opened new possibilities for ‘benchtop’ whole-genome sequencing. The Oxford Nanopore Technologies MinION is a portable device that uses nanopore technology that can directly sequence DNA molecules. MinION single molecule long sequence reads are well suited for de novo assembly of complex genomes as they facilitate the construction of highly contiguous physical genome maps obviating the need for labor-intensive physical genome mapping. Long sequence reads can also be used to delineate complex chromosomal rearrangements, such as those that occur in tumor cells, that can confound analysis using short reads. Here, we assessed MinION long-read-derived sequences for feasibility concerning: (1) the de novo assembly of a large complex genome, and (2) the elucidation of complex rearrangements. The genomes of two Caenorhabditis elegans strains, a wild-type strain and a strain containing two complex rearrangements, were sequenced with MinION. Up to 42-fold coverage was obtained from a single flow cell, and the best pooled data assembly produced a highly contiguous wild-type C. elegans genome containing 48 contigs (N50 contig length = 3.99 Mb) covering >99% of the 100,286,401-base reference genome. Further, the MinION-derived genome assembly expanded the C. elegans reference genome by >2 Mb due to a more accurate determination of repetitive sequence elements and assembled the complete genomes of two co-extracted bacteria. MinION long-read sequence data also facilitated the elucidation of complex rearrangements in a mutagenized strain. The sequence accuracy of the MinION long-read contigs (∼98%) was improved using Illumina-derived sequence data to polish the final genome assembly to 99.8% nucleotide accuracy when compared to the reference assembly.