scholarly journals Towards a better understanding of the low recall of insertion variants with short-read based variant callers

2020 ◽  
Author(s):  
Wesley Delage ◽  
Julien Thevenon ◽  
Claire Lemaitre
Keyword(s):  
Gold Standard ◽  
Insertion Site ◽  
Structural Variants ◽  
Genomic Context ◽  
Breakpoint Junction ◽  
Short Read ◽  
Depth Analysis ◽  
Long Read ◽  
The Impact

AbstractSince 2009, numerous tools have been developed to detect structural variants (SVs) using short read technologies. Insertions >50 bp are one of the hardest type to discover and are drastically underrepresented in gold standard variant callsets. The advent of long read technologies has completely changed the situation. In 2019, two independent cross technologies studies have published the most complete variant callsets with sequence resolved insertions in human individuals. Among the reported insertions, only 17 to 37% could be discovered with short-read based tools. In this work, we performed an in-depth analysis of these unprecedented insertion callsets in order to investigate the causes of such failures. We have first established a precise classification of insertion variants according to four layers of characterization: the nature and size of the inserted sequence, the genomic context of the insertion site and the breakpoint junction complexity. Because these levels are intertwined, we then used simulations to characterize the impact of each complexity factor on the recall of several SV callers. Simulations showed that the most impacting factor was the insertion type rather than the genomic context, with various difficulties being handled differently among the tested SV callers, and they highlighted the lack of sequence resolution for most insertion calls. Our results explain the low recall by pointing out several difficulty factors among the observed insertion features and provide avenues for improving SV caller algorithms and their [email protected]

scholarly journals Towards a better understanding of the low recall of insertion variants with short-read based variant callers

BMC Genomics ◽  
2020 ◽  
Vol 21 (1) ◽  
Author(s):  
Wesley J. Delage ◽  
Julien Thevenon ◽  
Claire Lemaitre
Keyword(s):  
Tandem Repeats ◽  
Insertion Site ◽  
Mobile Element ◽  
Genomic Context ◽  
Short Read ◽  
Structural Variant ◽  
Depth Analysis ◽  
Long Read ◽  
Repeat Expansions ◽  
The Impact

Abstract Background Since 2009, numerous tools have been developed to detect structural variants using short read technologies. Insertions >50 bp are one of the hardest type to discover and are drastically underrepresented in gold standard variant callsets. The advent of long read technologies has completely changed the situation. In 2019, two independent cross technologies studies have published the most complete variant callsets with sequence resolved insertions in human individuals. Among the reported insertions, only 17 to 28% could be discovered with short-read based tools. Results In this work, we performed an in-depth analysis of these unprecedented insertion callsets in order to investigate the causes of such failures. We have first established a precise classification of insertion variants according to four layers of characterization: the nature and size of the inserted sequence, the genomic context of the insertion site and the breakpoint junction complexity. Because these levels are intertwined, we then used simulations to characterize the impact of each complexity factor on the recall of several structural variant callers. We showed that most reported insertions exhibited characteristics that may interfere with their discovery: 63% were tandem repeat expansions, 38% contained homology larger than 10 bp within their breakpoint junctions and 70% were located in simple repeats. Consequently, the recall of short-read based variant callers was significantly lower for such insertions (6% for tandem repeats vs 56% for mobile element insertions). Simulations showed that the most impacting factor was the insertion type rather than the genomic context, with various difficulties being handled differently among the tested structural variant callers, and they highlighted the lack of sequence resolution for most insertion calls. Conclusions Our results explain the low recall by pointing out several difficulty factors among the observed insertion features and provide avenues for improving SV caller algorithms and their combinations.

scholarly journals Genotyping structural variants in pangenome graphs using the vg toolkit

2019 ◽  
Author(s):  
Glenn Hickey ◽  
David Heller ◽  
Jean Monlong ◽  
Jonas A. Sibbesen ◽  
Jouni Sirén ◽  
...  
Keyword(s):  
De Novo ◽  
State Of The Art ◽  
Effective Means ◽  
Point Mutations ◽  
Structural Variants ◽  
Short Read ◽  
Yeast Strains ◽  
Sequencing Studies ◽  
Long Read

AbstractStructural variants (SVs) remain challenging to represent and study relative to point mutations despite their demonstrated importance. We show that variation graphs, as implemented in the vg toolkit, provide an effective means for leveraging SV catalogs for short-read SV genotyping experiments. We benchmarked vg against state-of-the-art SV genotypers using three sequence-resolved SV catalogs generated by recent long-read sequencing studies. In addition, we use assemblies from 12 yeast strains to show that graphs constructed directly from aligned de novo assemblies improve genotyping compared to graphs built from intermediate SV catalogs in the VCF format.

scholarly journals Rapid Mycobacterium tuberculosis spoligotyping from uncorrected long reads using Galru

2020 ◽  
Author(s):  
Andrew J. Page ◽  
Nabil-Fareed Alikhan ◽  
Michael Strinden ◽  
Thanh Le Viet ◽  
Timofey Skvortsov
Keyword(s):  
Mycobacterium Tuberculosis ◽  
State Of The Art ◽  
Sequence Data ◽  
Human Pathogen ◽  
Sequencing Data ◽  
Short Read ◽  
Short Read Sequencing ◽  
Long Reads ◽  
Long Read

AbstractSpoligotyping of Mycobacterium tuberculosis provides a subspecies classification of this major human pathogen. Spoligotypes can be predicted from short read genome sequencing data; however, no methods exist for long read sequence data such as from Nanopore or PacBio. We present a novel software package Galru, which can rapidly detect the spoligotype of a Mycobacterium tuberculosis sample from as little as a single uncorrected long read. It allows for near real-time spoligotyping from long read data as it is being sequenced, giving rapid sample typing. We compare it to the existing state of the art software and find it performs identically to the results obtained from short read sequencing data. Galru is freely available from https://github.com/quadram-institute-bioscience/galru under the GPLv3 open source licence.

scholarly journals Mapping and phasing of structural variation in patient genomes using nanopore sequencing

2017 ◽  
Author(s):  
Mircea Cretu Stancu ◽  
Markus J. van Roosmalen ◽  
Ivo Renkens ◽  
Marleen Nieboer ◽  
Sjors Middelkamp ◽  
...  
Keyword(s):  
Single Molecule ◽  
De Novo ◽  
Structural Variants ◽  
Human Genetic Disease ◽  
Structural Genomic ◽  
Short Read ◽  
Sequencing Technologies ◽  
Genome Wide ◽  
Long Read ◽  
Complex Structural

AbstractStructural genomic variants form a common type of genetic alteration underlying human genetic disease and phenotypic variation. Despite major improvements in genome sequencing technology and data analysis, the detection of structural variants still poses challenges, particularly when variants are of high complexity. Emerging long-read single-molecule sequencing technologies provide new opportunities for detection of structural variants. Here, we demonstrate sequencing of the genomes of two patients with congenital abnormalities using the ONT MinION at 11x and 16x mean coverage, respectively. We developed a bioinformatic pipeline - NanoSV - to efficiently map genomic structural variants (SVs) from the long-read data. We demonstrate that the nanopore data are superior to corresponding short-read data with regard to detection of de novo rearrangements originating from complex chromothripsis events in the patients. Additionally, genome-wide surveillance of SVs, revealed 3,253 (33%) novel variants that were missed in short-read data of the same sample, the majority of which are duplications < 200bp in size. Long sequencing reads enabled efficient phasing of genetic variations, allowing the construction of genome-wide maps of phased SVs and SNVs. We employed read-based phasing to show that all de novo chromothripsis breakpoints occurred on paternal chromosomes and we resolved the long-range structure of the chromothripsis. This work demonstrates the value of long-read sequencing for screening whole genomes of patients for complex structural variants.

scholarly journals Comparison of different technologies for the decipherment of the whole genome sequence of Campylobacter jejuni BfR-CA-14430

Gut Pathogens ◽  
2019 ◽  
Vol 11 (1) ◽  
Author(s):  
Lennard Epping ◽  
Julia C. Golz ◽  
Marie-Theres Knüver ◽  
Charlotte Huber ◽  
Andrea Thürmer ◽  
...  
Keyword(s):  
Campylobacter Jejuni ◽  
Genome Sequence ◽  
Bacterial Species ◽  
Illumina Miseq ◽  
Chicken Meat ◽  
Whole Genome ◽  
Short Read ◽  
Plasmid Sequence ◽  
Depth Analysis ◽  
Long Read

Abstract Background Campylobacter jejuni is a zoonotic pathogen that infects the human gut through the food chain mainly by consumption of undercooked chicken meat, raw chicken cross-contaminated ready-to-eat food or by raw milk. In the last decades, C. jejuni has increasingly become the most common bacterial cause for food-born infections in high income countries, costing public health systems billions of euros each year. Currently, different whole genome sequencing techniques such as short-read bridge amplification and long-read single molecule real-time sequencing techniques are applied for in-depth analysis of bacterial species, in particular, Illumina MiSeq, PacBio and MinION. Results In this study, we analyzed a recently isolated C. jejuni strain from chicken meat by short- and long-read data from Illumina, PacBio and MinION sequencing technologies. For comparability, this strain is used in the German PAC-CAMPY research consortium in several studies, including phenotypic analysis of biofilm formation, natural transformation and in vivo colonization models. The complete assembled genome sequence most likely consists of a chromosome of 1,645,980 bp covering 1665 coding sequences as well as a plasmid sequence with 41,772 bp that encodes for 46 genes. Multilocus sequence typing revealed that the strain belongs to the clonal complex CC-21 (ST-44) which is known to be involved in C. jejuni human infections, including outbreaks. Furthermore, we discovered resistance determinants and a point mutation in the DNA gyrase (gyrA) that render the bacterium resistant against ampicillin, tetracycline and (fluoro-)quinolones. Conclusion The comparison of Illumina MiSeq, PacBio and MinION sequencing and analyses with different assembly tools enabled us to reconstruct a complete chromosome as well as a circular plasmid sequence of the C. jejuni strain BfR-CA-14430. Illumina short-read sequencing in combination with either PacBio or MinION can substantially improve the quality of the complete chromosome and epichromosomal elements on the level of mismatches and insertions/deletions, depending on the assembly program used.

scholarly journals LinkedSV for detection of mosaic structural variants from linked-read exome and genome sequencing data

2018 ◽  
Author(s):  
Li Fang ◽  
Charlly Kao ◽  
Michael V Gonzalez ◽  
Fernanda A Mafra ◽  
Renata Pellegrino da Silva ◽  
...  
Keyword(s):  
Exome Sequencing ◽  
Read Depth ◽  
Structural Variants ◽  
Sequencing Data ◽  
High Coverage ◽  
Short Read ◽  
Short Read Sequencing ◽  
Sequencing Studies ◽  
Long Read ◽  
Local Assembly

AbstractLinked-read sequencing provides long-range information on short-read sequencing data by barcoding reads originating from the same DNA molecule, and can improve the detection and breakpoint identification for structural variants (SVs). We present LinkedSV for SV detection on linked-read sequencing data. LinkedSV considers barcode overlapping and enriched fragment endpoints as signals to detect large SVs, while it leverages read depth, paired-end signals and local assembly to detect small SVs. Benchmarking studies demonstrates that LinkedSV outperforms existing tools, especially on exome data and on somatic SVs with low variant allele frequencies. We demonstrate clinical cases where LinkedSV identifies disease causal SVs from linked-read exome sequencing data missed by conventional exome sequencing, and show examples where LinkedSV identifies SVs missed by high-coverage long-read sequencing. In summary, LinkedSV can detect SVs missed by conventional short-read and long-read sequencing approaches, and may resolve negative cases from clinical genome/exome sequencing studies.

scholarly journals Contig annotation tool CAT robustly classifies assembled metagenomic contigs and long sequences

2016 ◽  
Author(s):  
Diego D. Cambuy ◽  
Felipe H. Coutinho ◽  
Bas E. Dutilh
Keyword(s):  
Single Molecule ◽  
Dna Sequences ◽  
Taxonomic Classification ◽  
Annotation Tool ◽  
Single Molecule Sequencing ◽  
Short Read ◽  
Long Read ◽  
Micro Organisms ◽  
Taxonomic Annotation

AbstractIn modern-day metagenomics, there is an increasing need for robust taxonomic annotation of long DNA sequences from unknown micro-organisms. Long metagenomic sequences may be derived from assembly of short-read metagenomes, or from long-read single molecule sequencing. Here we introduce CAT, a pipeline for robust taxonomic classification of long DNA sequences. We show that CAT correctly classifies contigs at different taxonomic levels, even in simulated metagenomic datasets that are very distantly related from the sequences in the database. CAT is implemented in Python and the required scripts can be freely downloaded from Github.

scholarly journals NextSV: a meta-caller for structural variants from low-coverage long-read sequencing data

2016 ◽  
Author(s):  
Li Fang ◽  
Jiang Hu ◽  
Depeng Wang ◽  
Kai Wang
Keyword(s):  
Whole Genome ◽  
Ashkenazi Jewish ◽  
Structural Variants ◽  
Sequencing Data ◽  
Short Read ◽  
Short Read Sequencing ◽  
Human Genomes ◽  
Long Read ◽  
Personal Genomes ◽  
Low Coverage

AbstractBackgroundStructural variants (SVs) in human genomes are implicated in a variety of human diseases. Long-read sequencing delivers much longer read lengths than short-read sequencing and may greatly improve SV detection. However, due to the relatively high cost of long-read sequencing, it is unclear what coverage is needed and how to optimally use the aligners and SV callers.ResultsIn this study, we developed NextSV, a meta-caller to perform SV calling from low coverage long-read sequencing data. NextSV integrates three aligners and three SV callers and generates two integrated call sets (sensitive/stringent) for different analysis purposes. We evaluated SV calling performance of NextSV under different PacBio coverages on two personal genomes, NA12878 and HX1. Our results showed that, compared with running any single SV caller, NextSV stringent call set had higher precision and balanced accuracy (F1 score) while NextSV sensitive call set had a higher recall. At 10X coverage, the recall of NextSV sensitive call set was 93.5% to 94.1% for deletions and 87.9% to 93.2% for insertions, indicating that ~10X coverage might be an optimal coverage to use in practice, considering the balance between the sequencing costs and the recall rates. We further evaluated the Mendelian errors on an Ashkenazi Jewish trio dataset.ConclusionsOur results provide useful guidelines for SV detection from low coverage whole-genome PacBio data and we expect that NextSV will facilitate the analysis of SVs on long-read sequencing data.

scholarly journals LinkedSV for detection of mosaic structural variants from linked-read exome and genome sequencing data

2019 ◽  
Vol 10 (1) ◽  
Author(s):  
Li Fang ◽  
Charlly Kao ◽  
Michael V. Gonzalez ◽  
Fernanda A. Mafra ◽  
Renata Pellegrino da Silva ◽  
...  
Keyword(s):  
Exome Sequencing ◽  
Read Depth ◽  
Structural Variants ◽  
Sequencing Data ◽  
High Coverage ◽  
Short Read ◽  
Short Read Sequencing ◽  
Sequencing Studies ◽  
Long Read ◽  
Local Assembly

AbstractLinked-read sequencing provides long-range information on short-read sequencing data by barcoding reads originating from the same DNA molecule, and can improve detection and breakpoint identification for structural variants (SVs). Here we present LinkedSV for SV detection on linked-read sequencing data. LinkedSV considers barcode overlapping and enriched fragment endpoints as signals to detect large SVs, while it leverages read depth, paired-end signals and local assembly to detect small SVs. Benchmarking studies demonstrate that LinkedSV outperforms existing tools, especially on exome data and on somatic SVs with low variant allele frequencies. We demonstrate clinical cases where LinkedSV identifies disease-causal SVs from linked-read exome sequencing data missed by conventional exome sequencing, and show examples where LinkedSV identifies SVs missed by high-coverage long-read sequencing. In summary, LinkedSV can detect SVs missed by conventional short-read and long-read sequencing approaches, and may resolve negative cases from clinical genome/exome sequencing studies.

scholarly journals TT-Mars: Structural Variants Assessment Based on Haplotype-resolved Assemblies

2021 ◽  
Author(s):  
Jianzhi Yang ◽  
Mark Chaisson
Keyword(s):  
Gold Standard ◽  
Method Development ◽  
Point Estimate ◽  
Personal Genome ◽  
Structural Variants ◽  
High Quality ◽  
Alternative Approach ◽  
Long Read ◽  
Genome Assemblies ◽  
Validation Tool

AbstractVariant benchmarking is a critical component of method development and evaluating the accuracy of studies of genetic variation. Currently, the best approach to evaluate the accuracy of a callset is the comparison against a well curated gold standard. In repetitive regions of the genome it may be difficult to establish what is the truth for a call, for example when different alignment scoring metrics provide equally supported but different variant calls in on the same data. Here we provide an alternative approach, TT-Mars, that takes advantage of the recent production of high-quality haplotype-resolved genome assemblies by evaluating variant calls based on how well their call reflects the content of the assembly, rather than comparing calls themselves. We used TT-Mars to assess callsets from different SV discovery methods on multiple human genome samples and demonstrated that it is capable at accurately classifying true positive and false positive SVs. On the HG002 personal genome, TT-Mars recapitulates 96.0%-99.6% of the validations made using the Genome in a Bottle gold standard callset evaluated by truvari, and evaluates an additional 121-10,966 variants across different callsets. Furthermore, with a group of high-quality assemblies, TT-Mars can evaluate performance of SV calling algorithms as a distribution rather than a point estimate. We also compare TT-Mars against the long-read based validation tool, VaPoR, and when assembly-based variant calls produced by dipcall are used as a gold standard. Compared with VaPoR, TT-Mars analyzes more calls on a long read callset by assessing more short variant calls (< 100 bases), while requiring smaller input. Compared with validation using dipcall variants, TT-Mars analyzes 1,497-2,229 more calls on long read callsets and has favorable results when candidate calls are fragmented into multiple calls in alignments. TT-Mars is available at https://github.com/ChaissonLab/TT-Mars.git with accompanying assembly data and corresponding liftover files.

Sign in / Sign up

Export Citation Format

Share Document