Read Error Correction Research Articles

Natrix2 – Improved amplicon workflow with novel Oxford Nanopore Technologies support and enhancements in clustering, classification and taxonomic databases

Sequencing of amplified DNA is the first step towards the generation of Amplicon Sequence Variants (ASVs) or Operational Taxonomic Units (OTUs) for biodiversity assessment and comparative analyses of environmental communities and microbiomes. Notably, the rapid advancements in sequencing technologies have paved the way for the growing utilization of third-generation long-read approaches in recent years. These sequence data imply increasing read lengths, higher error rates, and altered sequencing chemistry. Likewise, methods for amplicon classification and reference databases have progressed, leading to the expansion of taxonomic application areas and higher classification accuracy. With Natrix, a user-friendly and reducible workflow solution, processing of prokaryotic and eukaryotic environmental Illumina sequences using 16S or 18S is possible. Here, we present an updated version of the pipeline, Natrix2, which incorporates VSEARCH as an alternative clustering method with better performance for 16S metabarcoding approaches and mothur for taxonomic classification on further databases, including PR2, UNITE and SILVA. Additionally, Natrix2 includes the handling of Nanopore reads, which entails initial error correction and refinement of reads using Medaka and Racon to subsequently determine their taxonomic classification.

Aligning distant sequences to graphs using long seed sketches.

Sequence-to-graph alignment is crucial for applications such as variant genotyping, read error correction, and genome assembly. We propose a novel seeding approach that relies on long inexact matches rather than short exact matches, and show that it yields a better time-accuracy trade-off in settings with up to a [Formula: see text] mutation rate. We use sketches of a subset of graph nodes, which are more robust to indels, and store them in a k-nearest neighbor index to avoid the curse of dimensionality. Our approach contrasts with existing methods and highlights the important role that sketching into vector space can play in bioinformatics applications. We show that our method scales to graphs with 1 billion nodes and has quasi-logarithmic query time for queries with an edit distance of [Formula: see text] For such queries, longer sketch-based seeds yield a [Formula: see text] increase in recall compared with exact seeds. Our approach can be incorporated into other aligners, providing a novel direction for sequence-to-graph alignment.

CARE 2.0: reducing false-positive sequencing error corrections using machine learning

BackgroundNext-generation sequencing pipelines often perform error correction as a preprocessing step to obtain cleaned input data. State-of-the-art error correction programs are able to reliably detect and correct the majority of sequencing errors. However, they also introduce new errors by making false-positive corrections. These correction mistakes can have negative impact on downstream analysis, such as k-mer statistics, de-novo assembly, and variant calling. This motivates the need for more precise error correction tools.ResultsWe present CARE 2.0, a context-aware read error correction tool based on multiple sequence alignment targeting Illumina datasets. In addition to a number of newly introduced optimizations its most significant change is the replacement of CARE 1.0’s hand-crafted correction conditions with a novel classifier based on random decision forests trained on Illumina data. This results in up to two orders-of-magnitude fewer false-positive corrections compared to other state-of-the-art error correction software. At the same time, CARE 2.0 is able to achieve high numbers of true-positive corrections comparable to its competitors. On a simulated full human dataset with 914M reads CARE 2.0 generates only 1.2M false positives (FPs) (and 801.4M true positives (TPs)) at a highly competitive runtime while the best corrections achieved by other state-of-the-art tools contain at least 3.9M FPs and at most 814.5M TPs. Better de-novo assembly and improved k-mer analysis show the applicability of CARE 2.0 to real-world data.ConclusionFalse-positive corrections can negatively influence down-stream analysis. The precision of CARE 2.0 greatly reduces the number of those corrections compared to other state-of-the-art programs including BFC, Karect, Musket, Bcool, SGA, and Lighter. Thus, higher-quality datasets are produced which improve k-mer analysis and de-novo assembly in real-world datasets which demonstrates the applicability of machine learning techniques in the context of sequencing read error correction. CARE 2.0 is written in C++/CUDA for Linux systems and can be run on the CPU as well as on CUDA-enabled GPUs. It is available at https://github.com/fkallen/CARE.

Long-read metagenomics retrieves complete single-contig bacterial genomes from canine feces

BackgroundLong-read sequencing in metagenomics facilitates the assembly of complete genomes out of complex microbial communities. These genomes include essential biologic information such as the ribosomal genes or the mobile genetic elements, which are usually missed with short-reads. We applied long-read metagenomics with Nanopore sequencing to retrieve high-quality metagenome-assembled genomes (HQ MAGs) from a dog fecal sample.ResultsWe used nanopore long-read metagenomics and frameshift aware correction on a canine fecal sample and retrieved eight single-contig HQ MAGs, which were > 90% complete with < 5% contamination, and contained most ribosomal genes and tRNAs. At the technical level, we demonstrated that a high-molecular-weight DNA extraction improved the metagenomics assembly contiguity, the recovery of the rRNA operons, and the retrieval of longer and circular contigs that are potential HQ MAGs. These HQ MAGs corresponded to Succinivibrio, Sutterella, Prevotellamassilia, Phascolarctobacterium, Catenibacterium, Blautia, and Enterococcus genera. Linking our results to previous gastrointestinal microbiome reports (metagenome or 16S rRNA-based), we found that some bacterial species on the gastrointestinal tract seem to be more canid-specific –Succinivibrio, Prevotellamassilia, Phascolarctobacterium, Blautia_A sp900541345–, whereas others are more broadly distributed among animal and human microbiomes –Sutterella, Catenibacterium, Enterococcus, and Blautia sp003287895. Sutterella HQ MAG is potentially the first reported genome assembly for Sutterella stercoricanis, as assigned by 16S rRNA gene similarity. Moreover, we show that long reads are essential to detect mobilome functions, usually missed in short-read MAGs.ConclusionsWe recovered eight single-contig HQ MAGs from canine feces of a healthy dog with nanopore long-reads. We also retrieved relevant biological insights from these specific bacterial species previously missed in public databases, such as complete ribosomal operons and mobilome functions. The high-molecular-weight DNA extraction improved the assembly’s contiguity, whereas the high-accuracy basecalling, the raw read error correction, the assembly polishing, and the frameshift correction reduced the insertion and deletion errors. Both experimental and analytical steps ensured the retrieval of complete bacterial genomes.

Ratatosk: hybrid error correction of long reads enables accurate variant calling and assembly

A major challenge to long read sequencing data is their high error rate of up to 15%. We present Ratatosk, a method to correct long reads with short read data. We demonstrate on 5 human genome trios that Ratatosk reduces the error rate of long reads 6-fold on average with a median error rate as low as 0.22 %. SNP calls in Ratatosk corrected reads are nearly 99 % accurate and indel calls accuracy is increased by up to 37 %. An assembly of Ratatosk corrected reads from an Ashkenazi individual yields a contig N50 of 45 Mbp and less misassemblies than a PacBio HiFi reads assembly.

CARE: context-aware sequencing read error correction.

Error correction is a fundamental pre-processing step in many Next-Generation Sequencing (NGS) pipelines, in particular for de novo genome assembly. However, existing error correction methods either suffer from high false-positive rates since they break reads into independent k-mers or do not scale efficiently to large amounts of sequencing reads and complex genomes. We present CARE-an alignment-based scalable error correction algorithm for Illumina data using the concept of minhashing. Minhashing allows for efficient similarity search within large sequencing read collections which enables fast computation of high-quality multiple alignments. Sequencing errors are corrected by detailed inspection of the corresponding alignments. Our performance evaluation shows that CARE generates significantly fewer false-positive corrections than state-of-the-art tools (Musket, SGA, BFC, Lighter, Bcool, Karect) while maintaining a competitive number of true positives. When used prior to assembly it can achieve superior de novo assembly results for a number of real datasets. CARE is also the first multiple sequence alignment-based error corrector that is able to process a human genome Illumina NGS dataset in only 4 h on a single workstation using GPU acceleration. CARE is open-source software written in C++ (CPU version) and in CUDA/C++ (GPU version). It is licensed under GPLv3 and can be downloaded at https://github.com/fkallen/CARE. Supplementary data are available at Bioinformatics online.

SPAligner: alignment of long diverged molecular sequences to assembly graphs

BackgroundGraph-based representation of genome assemblies has been recently used in different contexts — from improved reconstruction of plasmid sequences and refined analysis of metagenomic data to read error correction and reference-free haplotype reconstruction. While many of these applications heavily utilize the alignment of long nucleotide sequences to assembly graphs, first general-purpose software tools for finding such alignments have been released only recently and their deficiencies and limitations are yet to be discovered. Moreover, existing tools can not perform alignment of amino acid sequences, which could prove useful in various contexts — in particular the analysis of metagenomic sequencing data.ResultsIn this work we present a novel SPAligner (Saint-Petersburg Aligner) tool for aligning long diverged nucleotide and amino acid sequences to assembly graphs. We demonstrate that SPAligner is an efficient solution for mapping third generation sequencing reads onto assembly graphs of various complexity and also show how it can facilitate the identification of known genes in complex metagenomic datasets.ConclusionsOur work will facilitate accelerating the development of graph-based approaches in solving sequence to genome assembly alignment problem. SPAligner is implemented as a part of SPAdes tools library and is available on Github.

Large scale genome skimming from herbarium material for accurate plant identification and phylogenomics

BackgroundHerbaria are valuable sources of extensive curated plant material that are now accessible to genetic studies because of advances in high-throughput, next-generation sequencing methods. As an applied assessment of large-scale recovery of plastid and ribosomal genome sequences from herbarium material for plant identification and phylogenomics, we sequenced 672 samples covering 21 families, 142 genera and 530 named and proposed named species. We explored the impact of parameters such as sample age, DNA concentration and quality, read depth and fragment length on plastid assembly error. We also tested the efficacy of DNA sequence information for identifying plant samples using 45 specimens recently collected in the Pilbara.ResultsGenome skimming was effective at producing genomic information at large scale. Substantial sequence information on the chloroplast genome was obtained from 96.1% of samples, and complete or near-complete sequences of the nuclear ribosomal RNA gene repeat were obtained from 93.3% of samples. We were able to extract sequences for the core DNA barcode regions rbcL and matK from 96 to 93.3% of samples, respectively. Read quality and DNA fragment length had significant effects on sequencing outcomes and error correction of reads proved essential. Assembly problems were specific to certain taxa with low GC and high repeat content (Goodenia, Scaevola, Cyperus, Bulbostylis, Fimbristylis) suggesting biological rather than technical explanations. The structure of related genomes was needed to guide the assembly of repeats that exceeded the read length. DNA-based matching proved highly effective and showed that the efficacy for species identification declined in the order cpDNA >> rDNA > matK >> rbcL.ConclusionsWe showed that a large-scale approach to genome sequencing using herbarium specimens produces high-quality complete cpDNA and rDNA sequences as a source of data for DNA barcoding and phylogenomics.

Post Transplantation Measurable Residual Disease (MRD) Monitoring Using Next-Generation Sequencing Is Highly Predictive for Relapseafter Allogeneic Stem Cell Transplantation

Post Transplantation Measurable Residual Disease (MRD) Monitoring Using Next-Generation Sequencing Is Highly Predictive for Relapseafter Allogeneic Stem Cell Transplantation

Athena: Automated Tuning of k-mer based Genomic Error Correction Algorithms using Language Models

The performance of most error-correction (EC) algorithms that operate on genomics reads is dependent on the proper choice of its configuration parameters, such as the value of k in k-mer based techniques. In this work, we target the problem of finding the best values of these configuration parameters to optimize error correction and consequently improve genome assembly. We perform this in an adaptive manner, adapted to different datasets and to EC tools, due to the observation that different configuration parameters are optimal for different datasets, i.e., from different platforms and species, and vary with the EC algorithm being applied. We use language modeling techniques from the Natural Language Processing (NLP) domain in our algorithmic suite, Athena, to automatically tune the performance-sensitive configuration parameters. Through the use of N-Gram and Recurrent Neural Network (RNN) language modeling, we validate the intuition that the EC performance can be computed quantitatively and efficiently using the “perplexity” metric, repurposed from NLP. After training the language model, we show that the perplexity metric calculated from a sample of the test (or production) data has a strong negative correlation with the quality of error correction of erroneous NGS reads. Therefore, we use the perplexity metric to guide a hill climbing-based search, converging toward the best configuration parameter value. Our approach is suitable for both de novo and comparative sequencing (resequencing), eliminating the need for a reference genome to serve as the ground truth. We find that Athena can automatically find the optimal value of k with a very high accuracy for 7 real datasets and using 3 different k-mer based EC algorithms, Lighter, Blue, and Racer. The inverse relation between the perplexity metric and alignment rate exists under all our tested conditions—for real and synthetic datasets, for all kinds of sequencing errors (insertion, deletion, and substitution), and for high and low error rates. The absolute value of that correlation is at least 73%. In our experiments, the best value of k found by Athena achieves an alignment rate within 0.53% of the oracle best value of k found through brute force searching (i.e., scanning through the entire range of k values). Athena’s selected value of k lies within the top-3 best k values using N-Gram models and the top-5 best k values using RNN models With best parameter selection by Athena, the assembly quality (NG50) is improved by a Geometric Mean of 4.72X across the 7 real datasets.

De novo Nanopore read quality improvement using deep learning

BackgroundLong read sequencing technologies such as Oxford Nanopore can greatly decrease the complexity of de novo genome assembly and large structural variation identification. Currently Nanopore reads have high error rates, and the errors often cluster into low-quality segments within the reads. The limited sensitivity of existing read-based error correction methods can cause large-scale mis-assemblies in the assembled genomes, motivating further innovation in this area.ResultsHere we developed a Convolutional Neural Network (CNN) based method, called MiniScrub, for identification and subsequent “scrubbing” (removal) of low-quality Nanopore read segments to minimize their interference in downstream assembly process. MiniScrub first generates read-to-read overlaps via MiniMap2, then encodes the overlaps into images, and finally builds CNN models to predict low-quality segments. Applying MiniScrub to real world control datasets under several different parameters, we show that it robustly improves read quality, and improves read error correction in the metagenome setting. Compared to raw reads, de novo genome assembly with scrubbed reads produces many fewer mis-assemblies and large indel errors.ConclusionsMiniScrub is able to robustly improve read quality of Oxford Nanopore reads, especially in the metagenome setting, making it useful for downstream applications such as de novo assembly. We propose MiniScrub as a tool for preprocessing Nanopore reads for downstream analyses. MiniScrub is open-source software and is available at https://bitbucket.org/berkeleylab/jgi-miniscrub.

150P - Liquid biopsy assays using combined circulating tumour cells and circulating tumour DNA in the same patients for the diagnosis of primary lung cancer

BackgroundLung cancer is the most common cause of cancer-related mortality worldwide. Early diagnosis and surgery may be one of the most important strategy for the treatment of lung cancer patients. However, the procedure of lung cancer diagnosis from initial suspicion to final confirmation is not efficient due to low sensitivity of current diagnostic methods and difficulties involved in tumor tissue biopsy in lung cancer compared to other cancer types. Therefore, we investigated the feasibility of liquid biopsy using circulating tumor cells (CTCs) and circulating tumor DNA (ctDNA) analysis for the better diagnosis of lung cancer. MethodsIn this study, we blindly analyzed both CTCs and ctDNA in the blood samples derived from 109 patients including histology-proven and clinically suspicious lung cancer patients to test the sensitivity of liquid biopsy methods. CTC analysis was done by CytoGen’s liquid biopsy platform, in which viable CTCs were isolated by size-based filtration with gravity and validated by immunostaining of EpCAM or CK, excluding CD45 positive cells and enumerated by CytoGen’s cell imaging software. Lung cancer diagnosis was predicted by a cut-off, 2 ≥ CTC in 5ml peripheral blood. For ctDNA analysis, EDGC F-Can platform was used to analyze single nucleotide variations of very low variant allele frequencies by utilizing unique molecular indexes and a novel read error correction algorithm. For comparison, we also analyzed the levels of conventional tumor markers (CEA, cyfra21-1 and NSE) in the patient cohort. ResultsCompared to the diagnostic sensitivities of conventional tumor markers (CEA 29%; cyfra21-1 41%; NSE 39%), both assays of CTCs and ctDNA showed higher diagnostic sensitivity in predicting primary lung cancer (CTCs 67%; ctDNA 83%). When the assays of CTCs and ctDNA were combined for the diagnosis, the sensitivity was increased up to 98%. ConclusionsCollectively, this study suggests that combined CTCs and ctDNA assay would be useful for the diagnosis of primary lung cancer. Legal entity responsible for the studyCytogen, Inc., EDGC, Inc., Department of Medicine, Samsung Medical Center FundingHas not received any funding DisclosureAll authors have declared no conflicts of interest.

Abstract 1663: Analytical methods to identify tumor heterogeneity and rare subclones in single cell DNA sequencing data from targeted panels

Abstract Current bulk sequencing paradigms are inadequate to characterize complicated biological systems where somatic variation is buried in the landscape of populations. Single-cell DNA sequencing can simultaneously identify zygosity, rare alleles and determine whether mutations co-occur within the same cell. While high-throughput single-cell DNA analysis has been a recent innovation, it is essential to develop new capabilities for assessing genetic variation present in rare cells and to better understand the role that these cells play in the evolution of tumor progression. To address these challenges and enable the characterization of genetic diversity in cancer cell populations, we developed analytical methods to identify mutation signatures which define subclones present in a tumor population. Here we present a workflow for subclone identification using data generated on the Tapestri single-cell DNA platform. The analytical pipeline steps include obtaining raw reads from the sequencer, removing adapters, aligning and mapping the reads, calling individual cells, error correction of reads assigned to cells and identifying genetic variants within each cell. After filtering for high quality variants, we then filter for data completeness to ensure high quality data is used in downstream processing. The variant-cell matrix is then subjected to clustering to identify subclones. Top variants that define the signature of each subclone are also identified. To validate our methodology, we used data from two different targeted panels on different model systems with known truth mutations. Our pipeline shows the distinct clusters correlating with titration and cell line ratios. Cluster associated signature mutations were also identified. The pipeline can be used for multi sample analysis with time-series data from diagnosis to relapse or from primary site to metastasis to understand clonal diversity. These data demonstrate the utility of the Tapestri platform, the analytical pipeline, and associated data visualization capability. Our approach addresses key issues of identifying rare subpopulations of cells down to 0.1%, and transforms the ability to accurately characterize clonal heterogeneity in tumor samples. This high throughput method advances research efforts to improve patient stratification and therapy selection for various cancer indications. Citation Format: Manimozhi Manivannan, Sombeet Sahu, Shu Wang, Dong Kim, Niranjan Vissa, Jordan Wilheim, Adam Sciambi, Nianzhen Li, Robert Durruthy-Durruthy, Anup Parikh, Keith Jones. Analytical methods to identify tumor heterogeneity and rare subclones in single cell DNA sequencing data from targeted panels [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1663.

Bit-parallel sequence-to-graph alignment.

MotivationGraphs are commonly used to represent sets of sequences. Either edges or nodes can be labeled by sequences, so that each path in the graph spells a concatenated sequence. Examples include graphs to represent genome assemblies, such as string graphs and de Bruijn graphs, and graphs to represent a pan-genome and hence the genetic variation present in a population. Being able to align sequencing reads to such graphs is a key step for many analyses and its applications include genome assembly, read error correction and variant calling with respect to a variation graph.ResultsWe generalize two linear sequence-to-sequence algorithms to graphs: the Shift-And algorithm for exact matching and Myers’ bitvector algorithm for semi-global alignment. These linear algorithms are both based on processing w sequence characters with a constant number of operations, where w is the word size of the machine (commonly 64), and achieve a speedup of up to w over naive algorithms. For a graph with nodes and edges and a sequence of length m, our bitvector-based graph alignment algorithm reaches a worst case runtime of for acyclic graphs and for arbitrary cyclic graphs. We apply it to five different types of graphs and observe a speedup between 3-fold and 20-fold compared with a previous (asymptotically optimal) alignment algorithm.Availability and implementation https://github.com/maickrau/GraphAligner Supplementary information Supplementary data are available at Bioinformatics online.

IonHammer: Homopolymer-Space Hamming Clustering for IonTorrent Read Error Correction.

We describe fast and accurate algorithm for IonTorrent read error correction capable of significantly reducing the number of sequencing errors over the wide range of data sets. IonHammer is implemented in C++ and is freely available as part of the SPAdes genome assembler package.

TraRECo: a greedy approach based de novo transcriptome assembler with read error correction using consensus matrix

BackgroundThe challenges when developing a good de novo transcriptome assembler include how to deal with read errors and sequence repeats. Almost all de novo assemblers utilize a de Bruijn graph, with which complexity grows linearly with data size while suffering from errors and repeats. Although one can correct the errors by inspecting the topological structure of the graph, this is not an easy task when there are too many branches. Two research directions are to improve either the graph reliability or the path search precision, and in this study, we focused on the former.ResultsWe present TraRECo, a greedy approach to de novo assembly employing error-aware graph construction. In the proposed approach, we built contigs by direct read alignment within a distance margin and performed a junction search to construct splicing graphs. While doing so, a contig of length l was represented by a 4 × l matrix (called a consensus matrix), in which each element was the base count of the aligned reads so far. A representative sequence was obtained by taking the majority in each column of the consensus matrix to be used for further read alignment. Once the splicing graphs had been obtained, we used IsoLasso to find paths with a noticeable read depth. The experiments using real and simulated reads show that the method provided considerable improvement in sensitivity and moderately better performance when comparing sensitivity and precision. This was achieved by the error-aware graph construction using the consensus matrix, with which the reads having errors were made usable for the graph construction (otherwise, they might have been eventually discarded). This improved the quality of the coverage depth information used in the subsequent path search step and finally the reliability of the graph.ConclusionsDe novo assembly is mainly used to explore undiscovered isoforms and must be able to represent as many reads as possible in an efficient way. In this sense, TraRECo provides us with a potential alternative for improving graph reliability even though the computational burden is much higher than the single k-mer in the de Bruijn graph approach.

Clustering of circular consensus sequences: accurate error correction and assembly of single molecule real-time reads from multiplexed amplicon libraries

BackgroundTargeted resequencing with high-throughput sequencing (HTS) platforms can be used to efficiently interrogate the genomes of large numbers of individuals. A critical issue for research and applications using HTS data, especially from long-read platforms, is error in base calling arising from technological limits and bioinformatic algorithms. We found that the community standard long amplicon analysis (LAA) module from Pacific Biosciences is prone to substantial bioinformatic errors that raise concerns about findings based on this pipeline, prompting the need for a new method.ResultsA single molecule real-time (SMRT) sequencing-error correction and assembly pipeline, C3S-LAA, was developed for libraries of pooled amplicons. By uniquely leveraging the structure of SMRT sequence data (comprised of multiple low quality subreads from which higher quality circular consensus sequences are formed) to cluster raw reads, C3S-LAA produced accurate consensus sequences and assemblies of overlapping amplicons from single sample and multiplexed libraries. In contrast, despite read depths in excess of 100X per amplicon, the standard long amplicon analysis module from Pacific Biosciences generated unexpected numbers of amplicon sequences with substantial inaccuracies in the consensus sequences. A bootstrap analysis showed that the C3S-LAA pipeline per se was effective at removing bioinformatic sources of error, but in rare cases a read depth of nearly 400X was not sufficient to overcome minor but systematic errors inherent to amplification or sequencing.ConclusionsC3S-LAA uses a divide and conquer processing algorithm for SMRT amplicon-sequence data that generates accurate consensus sequences and local sequence assemblies. Solving the confounding bioinformatic source of error in LAA allowed for the identification of limited instances of errors due to DNA amplification or sequencing of homopolymeric nucleotide tracts. For research and development in genomics, C3S-LAA allows meaningful conclusions and biological inferences to be made from accurately polished sequence output.

Hercules: a profile HMM-based hybrid error correction algorithm for long reads.

Choosing whether to use second or third generation sequencing platforms can lead to trade-offs between accuracy and read length. Several types of studies require long and accurate reads. In such cases researchers often combine both technologies and the erroneous long reads are corrected using the short reads. Current approaches rely on various graph or alignment based techniques and do not take the error profile of the underlying technology into account. Efficient machine learning algorithms that address these shortcomings have the potential to achieve more accurate integration of these two technologies. We propose Hercules, the first machine learning-based long read error correction algorithm. Hercules models every long read as a profile Hidden Markov Model with respect to the underlying platform’s error profile. The algorithm learns a posterior transition/emission probability distribution for each long read to correct errors in these reads. We show on two DNA-seq BAC clones (CH17-157L1 and CH17-227A2) that Hercules-corrected reads have the highest mapping rate among all competing algorithms and have the highest accuracy when the breadth of coverage is high. On a large human CHM1 cell line WGS data set, Hercules is one of the few scalable algorithms; and among those, it achieves the highest accuracy.

Hybrid correction of highly noisy long reads using a variable-order de Bruijn graph.

The recent rise of long read sequencing technologies such as Pacific Biosciences and Oxford Nanopore allows to solve assembly problems for larger and more complex genomes than what allowed short reads technologies. However, these long reads are very noisy, reaching an error rate of around 10-15% for Pacific Biosciences, and up to 30% for Oxford Nanopore. The error correction problem has been tackled by either self-correcting the long reads, or using complementary short reads in a hybrid approach. However, even though sequencing technologies promise to lower the error rate of the long reads below 10%, it is still higher in practice, and correcting such noisy long reads remains an issue. We present HG-CoLoR, a hybrid error correction method that focuses on a seed-and-extend approach based on the alignment of the short reads to the long reads, followed by the traversal of a variable-order de Bruijn graph, built from the short reads. Our experiments show that HG-CoLoR manages to efficiently correct highly noisy long reads that display an error rate as high as 44%. When compared to other state-of-the-art long read error correction methods, our experiments also show that HG-CoLoR provides the best trade-off between runtime and quality of the results, and is the only method able to efficiently scale to eukaryotic genomes. HG-CoLoR is implemented is C++, supported on Linux platforms and freely available at https://github.com/morispi/HG-CoLoR. Supplementary data are available at Bioinformatics online.

Integrating long-range connectivity information into de Bruijn graphs.

MotivationThe de Bruijn graph is a simple and efficient data structure that is used in many areas of sequence analysis including genome assembly, read error correction and variant calling. The data structure has a single parameter k, is straightforward to implement and is tractable for large genomes with high sequencing depth. It also enables representation of multiple samples simultaneously to facilitate comparison. However, unlike the string graph, a de Bruijn graph does not retain long range information that is inherent in the read data. For this reason, applications that rely on de Bruijn graphs can produce sub-optimal results given their input data.ResultsWe present a novel assembly graph data structure: the Linked de Bruijn Graph (LdBG). Constructed by adding annotations on top of a de Bruijn graph, it stores long range connectivity information through the graph. We show that with error-free data it is possible to losslessly store and recover sequence from a Linked de Bruijn graph. With assembly simulations we demonstrate that the LdBG data structure outperforms both our de Bruijn graph and the String Graph Assembler (SGA). Finally we apply the LdBG to Klebsiella pneumoniae short read data to make large (12 kbp) variant calls, which we validate using PacBio sequencing data, and to characterize the genomic context of drug-resistance genes.Availability and implementationLinked de Bruijn Graphs and associated algorithms are implemented as part of McCortex, which is available under the MIT license at https://github.com/mcveanlab/mccortex.Supplementary information Supplementary data are available at Bioinformatics online.