Small Variant Calling Research Articles

Pisces: an accurate and versatile variant caller for somatic and germline next-generation sequencing data.

MotivationNext-generation sequencing technology is transitioning quickly from research labs to clinical settings. The diagnosis and treatment selection for many acquired and autosomal conditions necessitate a method for accurately detecting somatic and germline variants.ResultsWe have developed Pisces, a rapid, versatile and accurate small-variant calling suite designed for somatic and germline amplicon sequencing applications. Accuracy is achieved by four distinct modules, each incorporating a number of novel algorithmic strategies.Availability and implementationPisces is distributed under an open source license and can be downloaded from https://github.com/Illumina/Pisces. Pisces is available on the BaseSpace™ SequenceHub. It is distributed on Illumina sequencing platforms such as the MiSeq™ and is included in the Praxis™ Extended RAS Panel test which was recently approved by the FDA.Supplementary information Supplementary data are available at Bioinformatics online.

Determining Performance Metrics for Targeted Next-Generation Sequencing Panels Using Reference Materials

The National Institute of Standards and Technology has developed reference materials for five human genomes. DNA aliquots are available for purchase, and the data, analyses, and high-confidence small variant and homozygous reference calls are freely available on the web. These reference materials are useful for evaluating whole-genome sequencing methods and also can be used to benchmark targeted sequencing panels, which are used commonly in clinical settings. This article describes how to use the Genome in a Bottle samples to obtain performance metrics on any germline-targeted sequencing panel of interest, as well as the limitations of the reference materials. These materials are useful for understanding the limitations of, and optimizing, targeted sequencing panels and associated bioinformatics pipelines. Example figures are presented to illustrate ways to access the performance metrics of targeted sequencing panels, and a table of best practices is included.

Abstract 428: Supporting neoantigen identification for personalized cancer vaccines through analytical validation of an augmented content enhanced (ACE) exome

Abstract The identification of neoantigens has become a critical step in the development of neoantigen-based personalized cancer vaccines and other immunotherapy applications. Since neoantigens can be generated from tumor specific mutations in any expressed gene, the first step in identification of neoantigens typically involves deep exome and transcriptome sequencing on the tumor and exome sequencing of the matched normal. As personalized vaccines enter clinical trials with the potential for clinical use, there is a growing need for strong analytical validation of these platforms. To address this we have developed our ACE Exome (~200X) and Transcriptome platforms for neoantigen identification which utilitize an augmented exome approach designed to increase sensitivity for neoantigens in low complexity, traditionally hard to sequencing regions. To enable this platform for neoantigen based personalized cancer vaccines, we have performed a validation of both our ACE Exome (tumor and normal) and ACE transcriptome platforms for detecting DNA-based SNVs and Indels, as well as for RNA based small variant and fusion calls. These are variant types are especially important for neoantigen identification. In this abstract we describe the ACE Exome validation. We used 11 cancer cell lines and their matched normals to assess analytical sensitivity and limits of detection (LOD) for small variant (SNV and Indel) detection using our ACE exome and Tumor Normal bioinformatics pipeline. We identified a gold set of variants, 875 SNVs and 19 Indels that were previously validated in these 11 cell lines (COSMIC, CCLE and Sanger Sequencing confirmed variants). These gold set variants were used to calculate our analytical sensitivity (percent of gold variants detected across the 11 cell line pairs using our assay). To determine our LOD, we chose 3 of the 11 cancer cell lines and created 6 dilutions (5%, 10%, 20%, 30%, 50% and 80% tumor purity) with their matched normal. We then determined Positive Predictive Agreement (PPA, percent of pure cell line variants detected in a diluted samples) and False Discovery Rate (FDR, percent of erroneously detected variants in the diluted sample that were not detected in the pure cell lines) metrics for variants across different minor allelic frequencies (MAF) in the diluted samples. The ACE “Tumor Normal” Exome assay had a high sensitivity of 98% for SNVs and 95% for Indels. The assay also showed robust PPA (sensitivity) of 97% and FDR (specificity) of 2% for SNVs with MAF >= 10% and PPA of 87% and FDR of 3% for Indels with MAF >= 10%. We demonstrate that the ACE “Tumor Normal” Exome assay is highly accurate for identification of SNVs and Indels in cancer exomes. With high analytical sensitivity, PPA and low FDR we believe this assay provides augmented ability to detect cancer driver and potential neoantigen generating mutations across various tumor types. Citation Format: Ravi K. Alla, Jennifer Yen, Sean M. Boyle, Richard Chen. Supporting neoantigen identification for personalized cancer vaccines through analytical validation of an augmented content enhanced (ACE) exome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 428. doi:10.1158/1538-7445.AM2017-428

Abstract 546: Supporting neoantigen identification for personalized cancer vaccines trough analytical validation of an augmented content enhanced (ACE) transcriptome

Abstract The identification of neoantigens has become a critical step in the development of neoantigen-based personalized cancer vaccines and other immunotherapy applications. Since neoantigens can be generated from tumor specific mutations in any expressed gene, the first step in neoantigen identification typically involves deep sequencing of the tumor exome and transcriptome combined with exome sequencing of the matched normal. As personalized vaccines enter clinical trials, there is a growing need for strong analytical validation of these platforms. To address this, we have developed the ACE Exome (~200X) and Transcriptome platforms for neoantigen identification which harness an augmented exome approach specifically designed to increase sensitivity for neoantigens in low complexity, difficult to sequence regions. To enable this platform for neoantigen based personalized cancer vaccines, we have performed a validation of both our ACE Exome (tumor and normal) and ACE transcriptome platforms for detecting DNA-based SNVs and Indels, as well as for RNA based small variant and fusion calls, variant types that are especially important for neoantigen identification. In this abstract, we describe the ACE Transcriptome validation. To validate the ACE Transcriptome, we assessed analytical sensitivity (AS) and positive predictive accuracy (PPA). AS was calculated from a reference set of 894 SNVs and 19 indels across 11 tumor cancer cell lines with matched normals. The reference set was constructed based on previously observed variants in CCLE, COSMIC or had been validated by Sanger sequencing. PPA and limit of detection (LOD) were calculated in a series of tumor-normal dilutions representing allele frequencies ranging from 10% to 100%. For fusions, we further selected an additional 10 cell lines to test the detection of 16 previously characterized fusion events including several clinically actionable fusions (ALK and BCR-ABL1). We report an analytical sensitivity for SNVs of >99%, and a PPA of >96% for small variants with >=10% minor allele frequency (MAF). For fusion events, we report an analytical sensitivity of >99%, with the detection of all 16 fusion events supported by at least 5 reads. We demonstrate that our ACE Transcriptome assay and RNA cancer pipeline is a highly sensitive and robust platform for detecting small variants and fusions in the RNA. Citation Format: Jennifer L. Yen, Sean Boyle, Ravi Alla, Jason Harris, Martina Lefterova, Richard Chen. Supporting neoantigen identification for personalized cancer vaccines trough analytical validation of an augmented content enhanced (ACE) transcriptome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 546. doi:10.1158/1538-7445.AM2017-546

Abstract 3732: Analytical performance of TruSight® Tumor 170 on small nucleotide variations and gene amplifications using DNA from formalin-fixed, paraffin-embedded (FFPE) solid tumor samples

Abstract Expanding the paradigm of solid tumor profiling from single-gene testing to comprehensive panels presents many challenges. One such challenges is the ability of these panels to detect genetic alterations from FFPE samples, where the DNA is of low abundance and often heavily compromised. Despite these challenges, next-generation sequencing (NGS) offers the ability to assess multiple variants simultaneously in an ever-expanding list of relevant tumor genes. To that end, Illumina developed a comprehensive, hybrid capture-based NGS assay targeting 170 key cancer genes that is FFPE optimized. The assay consists of a DNA workflow for the identification of single and multiple nucleotide variants (SNVs, MNVs), small insertions and deletions (indels), gene amplifications, as well as a RNA workflow for the identification of splice variants and gene fusions. Following sequencing on the NextSeq® or HiSeq® instruments, the analytical pipeline initiates variant calling. The DNA aligner and variant callers were first optimized against the simulated read data from >40,000 COSMIC[1] mutations reported in the exons of the 170 genes. To reduce false positive variant calling due to systematic errors, each variant call was evaluated against its locus specific background error distribution. This distribution was compiled from a panel of FFPE normal samples and was also used to normalize against systematic bias in read coverage to increase the accuracy of amplification calling. Furthermore, gene amplification calling was improved by the addition of enhancer probes to the hybrid capture pool. The analytical sensitivity and specificity of TruSight® Tumor 170* was assessed on a large collection of FFPE samples and reference material. A panel of 72 cancer samples, including multiple tissue types, reference standards, and cell line and FFPE mixes were used to evaluate the limit of detection. The samples contained 533 SNVs, 80 indels including deletions up to 30 base pairs and insertions up to 31 base pairs, 4 MNVs, and 31 gene amplifications, characterized by orthogonal testing methods. Using 40 ng DNA input, detection sensitivity of the >1000 variants (including replicates) tested at variant allele frequencies down to ~5% was at 99.6%, while detection sensitivity of gene amplifications as low as 1.45x to 2.2x was at 98%. For limit of blank samples, a panel of 24 normal samples was used. Again using 40 ng DNA input, we show >99% specificity for small variant calling and >95% specificity for gene amplification calling. These data demonstrates the TruSight® Tumor 170 is able to detect multiple variant types within a single sample at low nucleic acid input, while exhibiting high analytical sensitivity and specificity for low allele fraction detection. [1] Forbes, et al. (2015) *For Research Use Only. Not for use in diagnostic procedures. Citation Format: Danny Chou, Xiao Chen, Austin Purdy, Li Teng, Byron Luo, Chen Zhao, Laurel Ball, Allan Castaneda, Katie Clark, Brian Crain, Anthony Daulo, Manh Do, Tingting Du, Sarah Dumm, Yonmee Han, Michael Havern, Chia-Ling Hsieh, Tingting Jiang, Suzanne Johansen, Scott Lang, Rachel Liang, Jennifer S. LoCoco, Jaime McLean, Yousef Nassiri, Jason Rostron, Jennifer Silhavy, June Snedecor, Natasha Talago, Kevin Wu, Clare Zlatkov, Ali Kuraishy, Karen Gutekunst, Sohela De Rozieres, Matthew Friedenberg, Han-Yu Chuang, Anne C. Jager. Analytical performance of TruSight® Tumor 170 on small nucleotide variations and gene amplifications using DNA from formalin-fixed, paraffin-embedded (FFPE) solid tumor samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3732. doi:10.1158/1538-7445.AM2017-3732

Nanopore sequencing data analysis: state of the art, applications and challenges.

The nanopore sequencing process is based on the transit of a DNA molecule through a nanoscopic pore, and since the 90s is considered as one of the most promising approaches to detect polymeric molecules. In 2014, Oxford Nanopore Technologies (ONT) launched a beta-testing program that supplied the scientific community with the first prototype of a nanopore sequencer: the MinION. Thanks to this program, several research groups had the opportunity to evaluate the performance of this novel instrument and develop novel computational approaches for analyzing this new generation of data. Despite the short period of time from the release of the MinION, a large number of algorithms and tools have been developed for base calling, data handling, read mapping, de novo assembly and variant discovery. Here, we face the main computational challenges related to the analysis of nanopore data, and we carry out a comprehensive and up-to-date survey of the algorithmic solutions adopted by the bioinformatic community comparing performance and reporting limits and advantages of using this new generation of sequences for genomic analyses. Our analyses demonstrate that the use of nanopore data dramatically improves the de novo assembly of genomes and allows for the exploration of structural variants with an unprecedented accuracy and resolution. However, despite the impressive improvements reached by ONT in the past 2 years, the use of these data for small-variant calling is still challenging, and at present, it needs to be coupled with complementary short sequences for mitigating the intrinsic biases of nanopore sequencing technology.

VarMatch: robust matching of small variant datasets using flexible scoring schemes.

Small variant calling is an important component of many analyses, and, in many instances, it is important to determine the set of variants which appear in multiple callsets. Variant matching is complicated by variants that have multiple equivalent representations. Normalization and decomposition algorithms have been proposed, but are not robust to different representation of complex variants. Variant matching is also usually done to maximize the number of matches, as opposed to other optimization criteria. We present the VarMatch algorithm for the variant matching problem. Our algorithm is based on a theoretical result which allows us to partition the input into smaller subproblems without sacrificing accuracy. VarMatch is robust to different representation of complex variants and is particularly effective in low complexity regions or those dense in variants. VarMatch is able to detect more matches than either the normalization or decomposition algorithms on tested datasets. It also implements different optimization criteria, such as edit distance, that can improve robustness to different variant representations. Finally, the VarMatch software provides summary statistics, annotations and visualizations that are useful for understanding callers' performance. VarMatch is freely available at: https://github.com/medvedevgroup/varmatch. chensun@cse.psu.edu or pashadag@cse.psu.edu. Supplementary data are available at Bioinformatics online.

Svclassify: a method to establish benchmark structural variant calls.

BackgroundThe human genome contains variants ranging in size from small single nucleotide polymorphisms (SNPs) to large structural variants (SVs). High-quality benchmark small variant calls for the pilot National Institute of Standards and Technology (NIST) Reference Material (NA12878) have been developed by the Genome in a Bottle Consortium, but no similar high-quality benchmark SV calls exist for this genome. Since SV callers output highly discordant results, we developed methods to combine multiple forms of evidence from multiple sequencing technologies to classify candidate SVs into likely true or false positives. Our method (svclassify) calculates annotations from one or more aligned bam files from many high-throughput sequencing technologies, and then builds a one-class model using these annotations to classify candidate SVs as likely true or false positives.ResultsWe first used pedigree analysis to develop a set of high-confidence breakpoint-resolved large deletions. We then used svclassify to cluster and classify these deletions as well as a set of high-confidence deletions from the 1000 Genomes Project and a set of breakpoint-resolved complex insertions from Spiral Genetics. We find that likely SVs cluster separately from likely non-SVs based on our annotations, and that the SVs cluster into different types of deletions. We then developed a supervised one-class classification method that uses a training set of random non-SV regions to determine whether candidate SVs have abnormal annotations different from most of the genome. To test this classification method, we use our pedigree-based breakpoint-resolved SVs, SVs validated by the 1000 Genomes Project, and assembly-based breakpoint-resolved insertions, along with semi-automated visualization using svviz.ConclusionsWe find that candidate SVs with high scores from multiple technologies have high concordance with PCR validation and an orthogonal consensus method MetaSV (99.7 % concordant), and candidate SVs with low scores are questionable. We distribute a set of 2676 high-confidence deletions and 68 high-confidence insertions with high svclassify scores from these call sets for benchmarking SV callers. We expect these methods to be particularly useful for establishing high-confidence SV calls for benchmark samples that have been characterized by multiple technologies.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-016-2366-2) contains supplementary material, which is available to authorized users.

Strelka: accurate somatic small-variant calling from sequenced tumor–normal sample pairs

Whole genome and exome sequencing of matched tumor-normal sample pairs is becoming routine in cancer research. The consequent increased demand for somatic variant analysis of paired samples requires methods specialized to model this problem so as to sensitively call variants at any practical level of tumor impurity. We describe Strelka, a method for somatic SNV and small indel detection from sequencing data of matched tumor-normal samples. The method uses a novel Bayesian approach which represents continuous allele frequencies for both tumor and normal samples, while leveraging the expected genotype structure of the normal. This is achieved by representing the normal sample as a mixture of germline variation with noise, and representing the tumor sample as a mixture of the normal sample with somatic variation. A natural consequence of the model structure is that sensitivity can be maintained at high tumor impurity without requiring purity estimates. We demonstrate that the method has superior accuracy and sensitivity on impure samples compared with approaches based on either diploid genotype likelihoods or general allele-frequency tests. The Strelka workflow source code is available at ftp://strelka@ftp.illumina.com/. csaunders@illumina.com