An ensemble strategy that significantly improves de novo assembly of microbial genomes from metagenomic next-generation sequencing data

Next-generation sequencing (NGS) approaches rapidly produce millions to billions of short reads, which allow pathogen detection and discovery in human clinical, animal and environmental samples. A major limitation of sequence homology-based identification for highly divergent microorganisms is the short length of reads generated by most highly parallel sequencing technologies. Short reads require a high level of sequence similarities to annotated genes to confidently predict gene function or homology. Such recognition of highly divergent homologues can be improved by reference-free (de novo) assembly of short overlapping sequence reads into larger contigs. We describe an ensemble strategy that integrates the sequential use of various de Bruijn graph and overlap-layout-consensus assemblers with a novel partitioned sub-assembly approach. We also proposed new quality metrics that are suitable for evaluating metagenome de novo assembly. We demonstrate that this new ensemble strategy tested using in silico spike-in, clinical and environmental NGS datasets achieved significantly better contigs than current approaches.     

Identification of Optimum Sequencing Depth Especially for De Novo Genome Assembly of Small Genomes Using Next Generation Sequencing Data

Next Generation Sequencing (NGS) is a disruptive technology that has found widespread acceptance in the life sciences research community. The high throughput and low cost of sequencing has encouraged researchers to undertake ambitious genomic projects, especially in de novo genome sequencing. Currently, NGS systems generate sequence data as short reads and de novo genome assembly using these short reads is computationally very intensive. Due to lower cost of sequencing and higher throughput, NGS systems now provide the ability to sequence genomes at high depth. However, currently no report is available highlighting the impact of high sequence depth on genome assembly using real data sets and multiple assembly algorithms. Recently, some studies have evaluated the impact of sequence coverage, error rate and average read length on genome assembly using multiple assembly algorithms, however, these evaluations were performed using simulated datasets. One limitation of using simulated datasets is that variables such as error rates, read length and coverage which are known to impact genome assembly are carefully controlled. Hence, this study was undertaken to identify the minimum depth of sequencing required for de novo assembly for different sized genomes using graph based assembly algorithms and real datasets. Illumina reads for E.coli (4.6 MB) S.kudriavzevii (11.18 MB) and C.elegans (100 MB) were assembled using SOAPdenovo, Velvet, ABySS, Meraculous and IDBA-UD. Our analysis shows that 50X is the optimum read depth for assembling these genomes using all assemblers except Meraculous which requires 100X read depth. Moreover, our analysis shows that de novo assembly from 50X read data requires only 6–40 GB RAM depending on the genome size and assembly algorithm used. We believe that this information can be extremely valuable for researchers in designing experiments and multiplexing which will enable optimum utilization of sequencing as well as analysis resources.     

SMRT sequencing only de novo assembly of the sugar beet (Beta vulgaris) chloroplast genome

Background

Third generation sequencing methods, like SMRT (Single Molecule, Real-Time) sequencing developed by Pacific Biosciences, offer much longer read length in comparison to Next Generation Sequencing (NGS) methods. Hence, they are well suited for de novo- or re-sequencing projects. Sequences generated for these purposes will not only contain reads originating from the nuclear genome, but also a significant amount of reads originating from the organelles of the target organism. These reads are usually discarded but they can also be used for an assembly of organellar replicons. The long read length supports resolution of repetitive regions and repeats within the organelles genome which might be problematic when just using short read data. Additionally, SMRT sequencing is less influenced by GC rich areas and by long stretches of the same base.

Results

We describe a workflow for a de novo assembly of the sugar beet (Beta vulgaris ssp. vulgaris) chloroplast genome sequence only based on data originating from a SMRT sequencing dataset targeted on its nuclear genome. We show that the data obtained from such an experiment are sufficient to create a high quality assembly with a higher reliability than assemblies derived from e.g. Illumina reads only. The chloroplast genome is especially challenging for de novo assembling as it contains two large inverted repeat (IR) regions. We also describe some limitations that still apply even though long reads are used for the assembly.

Conclusions

SMRT sequencing reads extracted from a dataset created for nuclear genome (re)sequencing can be used to obtain a high quality de novo assembly of the chloroplast of the sequenced organism. Even with a relatively small overall coverage for the nuclear genome it is possible to collect more than enough reads to generate a high quality assembly that outperforms short read based assemblies. However, even with long reads it is not always possible to clarify the order of elements of a chloroplast genome sequence reliantly which we could demonstrate with Fosmid End Sequences (FES) generated with Sanger technology. Nevertheless, this limitation also applies to short read sequencing data but is reached in this case at a much earlier stage during finishing.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-015-0726-6) contains supplementary material, which is available to authorized users.     

Rapid hybrid de novo assembly of a microbial genome using only short reads: Corynebacterium pseudotuberculosis I19 as a case study

Due to the advent of the so-called Next-Generation Sequencing (NGS) technologies the amount of monetary and temporal resources for whole-genome sequencing has been reduced by several orders of magnitude. Sequence reads can be assembled either by anchoring them directly onto an available reference genome (classical reference assembly), or can be concatenated by overlap (de novo assembly). The latter strategy is preferable because it tends to maintain the architecture of the genome sequence the however, depending on the NGS platform used, the shortness of read lengths cause tremendous problems the in the subsequent genome assembly phase, impeding closing of the entire genome sequence. To address the problem, we developed a multi-pronged hybrid de novo strategy combining De Bruijn graph and Overlap-Layout-Consensus methods, which was used to assemble from short reads the entire genome of Corynebacterium pseudotuberculosis strain I19, a bacterium with immense importance in veterinary medicine that causes Caseous Lymphadenitis in ruminants, principally ovines and caprines. Briefly, contigs were assembled de novo from the short reads and were only oriented using a reference genome by anchoring. Remaining gaps were closed using iterative anchoring of short reads by craning to gap flanks. Finally, we compare the genome sequence assembled using our hybrid strategy to a classical reference assembly using the same data as input and show that with the availability of a reference genome, it pays off to use the hybrid de novo strategy, rather than a classical reference assembly, because more genome sequences are preserved using the former.     

Improving ancient DNA read mapping against modern reference genomes

ABSTRACT: BACKGROUND: Next-Generation Sequencing has revolutionized our approach to ancient DNA (aDNA) research, by providing complete genomic sequences of ancient individuals and extinct species. However, the recovery of genetic material from long-dead organisms is still complicated by a number of issues, including post-mortem DNA damage and high levels of environmental contamination. Together with error profiles specific to the type of sequencing platforms used, these specificities could limit our ability to map sequencing reads against modern reference genomes and therefore limit our ability to identify endogenous ancient reads, reducing the efficiency of shotgun sequencing aDNA. RESULTS: In this study, we compare different computational methods for improving the accuracy and sensitivity of aDNA sequence identification, based on shotgun sequencing reads recovered from Pleistocene horse extracts using Illumina GAIIx and Helicos Heliscope platforms. We show that the performance of the Burrows Wheeler Aligner (BWA), that has been developed for mapping of undamaged sequencing reads using platforms with low rates of indel-types of sequencing errors, can be employed at acceptable run-times by modifying default parameters in a platform-specific manner. We also examine if trimming likely damaged positions at read ends can increase the recovery of genuine aDNA fragments and if accurate identification of human contamination can be achieved using a strategy previously suggested based on best hit filtering. We show that combining our different mapping and filtering approaches can increase the number of high-quality endogenous hits recovered by up to 33%. CONCLUSIONS: We have shown that Illumina and Helicos sequences recovered from aDNA extracts could not be aligned to modern reference genomes with the same efficiency unless mapping parameters are optimized for the specific types of errors generated by these platforms and by post-mortem DNA damage. Our findings have important implications for future aDNA research, as we define mapping guidelines that improve our ability to identify genuine aDNA sequences, which in turn could improve the genotyping accuracy of ancient specimens. Our framework provides a significant improvement to the standard procedures used for characterizing ancient genomes, which is challenged by contamination and often low amounts of DNA material.     

Is the whole greater than the sum of its parts? De novo assembly strategies for bacterial genomes based on paired-end sequencing

Background

Whole genome sequence construction is becoming increasingly feasible because of advances in next generation sequencing (NGS), including increasing throughput and read length. By simply overlapping paired-end reads, we can obtain longer reads with higher accuracy, which can facilitate the assembly process. However, the influences of different library sizes and assembly methods on paired-end sequencing-based de novo assembly remain poorly understood.

Results

We used 250 bp Illumina Miseq paired-end reads of different library sizes generated from genomic DNA from Escherichia coli DH1 and Streptococcus parasanguinis FW213 to compare the assembly results of different library sizes and assembly approaches. Our data indicate that overlapping paired-end reads can increase read accuracy but sometimes cause insertion or deletions. Regarding genome assembly, merged reads only outcompete original paired-end reads when coverage depth is low, and larger libraries tend to yield better assembly results. These results imply that distance information is the most critical factor during assembly. Our results also indicate that when depth is sufficiently high, assembly from subsets can sometimes produce better results.

Conclusions

In summary, this study provides systematic evaluations of de novo assembly from paired end sequencing data. Among the assembly strategies, we find that overlapping paired-end reads is not always beneficial for bacteria genome assembly and should be avoided or used with caution especially for genomes containing high fraction of repetitive sequences. Because increasing numbers of projects aim at bacteria genome sequencing, our study provides valuable suggestions for the field of genomic sequence construction.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1859-8) contains supplementary material, which is available to authorized users.     

DecGPU: distributed error correction on massively parallel graphics processing units using CUDA and MPI

Background  

Next-generation sequencing technologies have led to the high-throughput production of sequence data (reads) at low cost. However, these reads are significantly shorter and more error-prone than conventional Sanger shotgun reads. This poses a challenge for the de novo assembly in terms of assembly quality and scalability for large-scale short read datasets.     

Reevaluating Assembly Evaluations with Feature Response Curves: GAGE and Assemblathons

In just the last decade, a multitude of bio-technologies and software pipelines have emerged to revolutionize genomics. To further their central goal, they aim to accelerate and improve the quality of de novo whole-genome assembly starting from short DNA sequences/reads. However, the performance of each of these tools is contingent on the length and quality of the sequencing data, the structure and complexity of the genome sequence, and the resolution and quality of long-range information. Furthermore, in the absence of any metric that captures the most fundamental “features” of a high-quality assembly, there is no obvious recipe for users to select the most desirable assembler/assembly. This situation has prompted the scientific community to rely on crowd-sourcing through international competitions, such as Assemblathons or GAGE, with the intention of identifying the best assembler(s) and their features. Somewhat circuitously, the only available approach to gauge de novo assemblies and assemblers relies solely on the availability of a high-quality fully assembled reference genome sequence. Still worse, reference-guided evaluations are often both difficult to analyze, leading to conclusions that are difficult to interpret. In this paper, we circumvent many of these issues by relying upon a tool, dubbed , which is capable of evaluating de novo assemblies from the read-layouts even when no reference exists. We extend the FRCurve approach to cases where lay-out information may have been obscured, as is true in many deBruijn-graph-based algorithms. As a by-product, FRCurve now expands its applicability to a much wider class of assemblers – thus, identifying higher-quality members of this group, their inter-relations as well as sensitivity to carefully selected features, with or without the support of a reference sequence or layout for the reads. The paper concludes by reevaluating several recently conducted assembly competitions and the datasets that have resulted from them.     

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.     

Evaluation and Validation of Assembling Corrected PacBio Long Reads for Microbial Genome Completion via Hybrid Approaches

Despite the ever-increasing output of next-generation sequencing data along with developing assemblers, dozens to hundreds of gaps still exist in de novo microbial assemblies due to uneven coverage and large genomic repeats. Third-generation single-molecule, real-time (SMRT) sequencing technology avoids amplification artifacts and generates kilobase-long reads with the potential to complete microbial genome assembly. However, due to the low accuracy (~85%) of third-generation sequences, a considerable amount of long reads (>50X) are required for self-correction and for subsequent de novo assembly. Recently-developed hybrid approaches, using next-generation sequencing data and as few as 5X long reads, have been proposed to improve the completeness of microbial assembly. In this study we have evaluated the contemporary hybrid approaches and demonstrated that assembling corrected long reads (by runCA) produced the best assembly compared to long-read scaffolding (e.g., AHA, Cerulean and SSPACE-LongRead) and gap-filling (SPAdes). For generating corrected long reads, we further examined long-read correction tools, such as ECTools, LSC, LoRDEC, PBcR pipeline and proovread. We have demonstrated that three microbial genomes including Escherichia coli K12 MG1655, Meiothermus ruber DSM1279 and Pdeobacter heparinus DSM2366 were successfully hybrid assembled by runCA into near-perfect assemblies using ECTools-corrected long reads. In addition, we developed a tool, Patch, which implements corrected long reads and pre-assembled contigs as inputs, to enhance microbial genome assemblies. With the additional 20X long reads, short reads of S. cerevisiae W303 were hybrid assembled into 115 contigs using the verified strategy, ECTools + runCA. Patch was subsequently applied to upgrade the assembly to a 35-contig draft genome. Our evaluation of the hybrid approaches shows that assembling the ECTools-corrected long reads via runCA generates near complete microbial genomes, suggesting that genome assembly could benefit from re-analyzing the available hybrid datasets that were not assembled in an optimal fashion.     

Protein Domain Analysis of Genomic Sequence Data Reveals Regulation of LRR Related Domains in Plant Transpiration in Ficus

Predicting protein domains is essential for understanding a protein’s function at the molecular level. However, up till now, there has been no direct and straightforward method for predicting protein domains in species without a reference genome sequence. In this study, we developed a functionality with a set of programs that can predict protein domains directly from genomic sequence data without a reference genome. Using whole genome sequence data, the programming functionality mainly comprised DNA assembly in combination with next-generation sequencing (NGS) assembly methods and traditional methods, peptide prediction and protein domain prediction. The proposed new functionality avoids problems associated with de novo assembly due to micro reads and small single repeats. Furthermore, we applied our functionality for the prediction of leucine rich repeat (LRR) domains in four species of Ficus with no reference genome, based on NGS genomic data. We found that the LRRNT_2 and LRR_8 domains are related to plant transpiration efficiency, as indicated by the stomata index, in the four species of Ficus. The programming functionality established in this study provides new insights for protein domain prediction, which is particularly timely in the current age of NGS data expansion.     

MetaVelvet: an extension of Velvet assembler to de novo metagenome assembly from short sequence reads

An important step in ‘metagenomics’ analysis is the assembly of multiple genomes from mixed sequence reads of multiple species in a microbial community. Most conventional pipelines use a single-genome assembler with carefully optimized parameters. A limitation of a single-genome assembler for de novo metagenome assembly is that sequences of highly abundant species are likely misidentified as repeats in a single genome, resulting in a number of small fragmented scaffolds. We extended a single-genome assembler for short reads, known as ‘Velvet’, to metagenome assembly, which we called ‘MetaVelvet’, for mixed short reads of multiple species. Our fundamental concept was to first decompose a de Bruijn graph constructed from mixed short reads into individual sub-graphs, and second, to build scaffolds based on each decomposed de Bruijn sub-graph as an isolate species genome. We made use of two features, the coverage (abundance) difference and graph connectivity, for the decomposition of the de Bruijn graph. For simulated datasets, MetaVelvet succeeded in generating significantly higher N50 scores than any single-genome assemblers. MetaVelvet also reconstructed relatively low-coverage genome sequences as scaffolds. On real datasets of human gut microbial read data, MetaVelvet produced longer scaffolds and increased the number of predicted genes.     

De novo ChIP-seq analysis

Methods for the analysis of chromatin immunoprecipitation sequencing (ChIP-seq) data start by aligning the short reads to a reference genome. While often successful, they are not appropriate for cases where a reference genome is not available. Here we develop methods for de novo analysis of ChIP-seq data. Our methods combine de novo assembly with statistical tests enabling motif discovery without the use of a reference genome. We validate the performance of our method using human and mouse data. Analysis of fly data indicates that our method outperforms alignment based methods that utilize closely related species.

Electronic supplementary material

The online version of this article (doi:10.1186/s13059-015-0756-4) contains supplementary material, which is available to authorized users.     

Microarray meta-analysis database (M2DB): a uniformly pre-processed, quality controlled, and manually curated human clinical microarray database

Background

With the rapid expansion of DNA sequencing databases, it is now feasible to identify relevant information from prior sequencing projects and completed genomes and apply it to de novo sequencing of new organisms. As an example, this paper demonstrates how such extra information can be used to improve de novo assemblies by augmenting the overlapping step. Finding all pairs of overlapping reads is a key task in many genome assemblers, and to this end, highly efficient algorithms have been developed to find alignments in large collections of sequences. It is well known that due to repeated sequences, many aligned pairs of reads nevertheless do not overlap. But no overlapping algorithm to date takes a rigorous approach to separating aligned but non-overlapping read pairs from true overlaps.

Results

We present an approach that extends the Minimus assembler by a data driven step to classify overlaps as true or false prior to contig construction. We trained several different classification models within the Weka framework using various statistics derived from overlaps of reads available from prior sequencing projects. These statistics included percent mismatch and k-mer frequencies within the overlaps as well as a comparative genomics score derived from mapping reads to multiple reference genomes. We show that in real whole-genome sequencing data from the E. coli and S. aureus genomes, by providing a curated set of overlaps to the contigging phase of the assembler, we nearly doubled the median contig length (N50) without sacrificing coverage of the genome or increasing the number of mis-assemblies.

Conclusions

Machine learning methods that use comparative and non-comparative features to classify overlaps as true or false can be used to improve the quality of a sequence assembly.     

A biologist's guide to de novo genome assembly using next-generation sequence data: A test with fungal genomes

We offer a guide to de novo genome assembly¹ using sequence data generated by the Illumina platform for biologists working with fungi or other organisms whose genomes are less than 100 Mb in size. The guide requires no familiarity with sequencing assembly technology or associated computer programs. It defines commonly used terms in genome sequencing and assembly; provides examples of assembling short-read genome sequence data for four strains of the fungus Grosmannia clavigera using four assembly programs; gives examples of protocols and software; and presents a commented flowchart that extends from DNA preparation for submission to a sequencing center, through to processing and assembly of the raw sequence reads using freely available operating systems and software.