Single Nucleotide Variant Calling Research Articles

MiniSNV: accurate and fast single nucleotide variant calling from nanopore sequencing data.

Nanopore sequence technology has demonstrated a longer read length and enabled to potentially address the limitations of short-read sequencing including long-range haplotype phasing and accurate variant calling. However, there is still room for improvement in terms of the performance of single nucleotide variant (SNV) identification and computing resource usage for the state-of-the-art approaches. In this work, we introduce miniSNV, a lightweight SNV calling algorithm that simultaneously achieves high performance and yield. miniSNV utilizes known common variants in populations as variation backgrounds and leverages read pileup, read-based phasing, and consensus generation to identify and genotype SNVs for Oxford Nanopore Technologies (ONT) long reads. Benchmarks on real and simulated ONT data under various error profiles demonstrate that miniSNV has superior sensitivity and comparable accuracy on SNV detection and runs faster with outstanding scalability and lower memory than most state-of-the-art variant callers. miniSNV is available from https://github.com/CuiMiao-HIT/miniSNV.

Identifying cancer cells from calling single-nucleotide variants in scRNA-seq data.

Single-cell RNA sequencing (scRNA-seq) data are widely used to study cancer cell states and their heterogeneity. However, the tumour microenvironment is usually a mixture of healthy and cancerous cells and it can be difficult to fully separate these two populations based on transcriptomics alone. If available, somatic single-nucleotide variants (SNVs) observed in the scRNA-seq data could be used to identify the cancer population and match that information with the single cells' expression profile. However, calling somatic SNVs in scRNA-seq data is a challenging task, as most variants seen in the short-read data are not somatic, but can instead be germline variants, RNA edits or transcription, sequencing, or processing errors. In addition, only variants present in actively transcribed regions for each individual cell will be seen in the data. To address these challenges, we develop CCLONE (Cancer Cell Labelling On Noisy Expression), an interpretable tool adapted to handle the uncertainty and sparsity of SNVs called from scRNA-seq data. CCLONE jointly identifies cancer clonal populations, and their associated variants. We apply CCLONE on two acute myeloid leukaemia datasets and one lung adenocarcinoma dataset and show that CCLONE captures both genetic clones and somatic events for multiple patients. These results show how CCLONE can be used to gather insight into the course of the disease and the origin of cancer cells in scRNA-seq data. Source code is available at github.com/HaghverdiLab/CCLONE.

Development and application of an ultra-sensitive ctDNA mutation profiling assay in monitoring therapy response and drug resistance.

264 Background: Next-generation sequencing (NGS) assays offer efficient high-throughput profiling methods for analyzing circulating tumor DNA (ctDNA) in blood or other liquid biopsy specimens. The ctDNA mutant allele fractions of patients after treatment is often very low and below the detection limit of regular cfDNA assays. Here we report the development and clinical application of a CLIA-certified ultra-sensitive NGS cfDNA assay in detecting the low-frequency ctDNA mutations and monitoring therapy response and drug resistance in patients with metastatic gastroesophageal adenocarcinoma under combined blockade treatments. Methods: PredicineCARE ULTRA cfDNA assay with 100,000X sequencing depth was validated under CLIA settings. Limit of blank (LoB) and limit of detection (LoD) were established by profiling 38 healthy donor plasma samples, 23 contrived human cfDNAs samples from patients with various cancers, and 30 replicates of well-established ctDNA reference standards. In silico down-sampling of FASTQ files was carried out to test the sensitivity of this assay with a lower coverage (50,000X). This assay was then adopted for longitudinal sequencing of 57 serial plasmas from 6 patients with HER2+ gastroesophageal adenocarcinomas (GEC) who underwent combined treatments (capecitabine, oxaliplatin, bevacizumab and trastuzumab) at different time points (baseline, response, and progression). Results: PredicineCARE ULTRA cfDNA assay exhibited exceptional sensitivity in calling single-nucleotide variants and indels, with a 98.0% detection rate at a LoD of 0.075% mutant allele frequency (MAF) for 60 ng cfDNA. This is significantly lower than the LoD (0.25% MAF) of a standard PredicineCARE cfDNA assay. The ultra-sensitive assay also displayed decent analytical specificity (>99.99%) and precision (intra-assay 100%; inter-assay 99.41%). In 6 GEC patients’ longitudinal samples, PredicineCARE ULTRA cfDNA assay detected 36 additional somatic mutations with the MAFs spanning in a range from 0.024% to 0.1% beyond 124 baseline mutations identified using standard PredicineCARE. In two representative cases presented here, all baseline mutations (genes TP53 and PIK3CA etc.) were consistently reported in the follow-up time points with a noticeable MAF reduction (lowest MAFs from 0.046% to 0.024%) during response. Subsequent elevations were observed in more genes, including low-frequency mutations (KRAS and ERBB2) not detected at baseline or only detected in PredicineCARE ULTRA. Conclusions: This study demonstrated the promise and clinical utility of the ultra-sensitive cfDNA molecular profiling at 100,000X sequencing depth in treatment selection, therapy monitoring, and drug resistance studies.

Empirical Bayes single nucleotide variant-calling for next-generation sequencing data

One of the fundamental computational problems in cancer genomics is the identification of single nucleotide variants (SNVs) from DNA sequencing data. Many statistical models and software implementations for SNV calling have been developed in the literature, yet, they still disagree widely on real datasets. Based on an empirical Bayesian approach, we introduce a local false discovery rate (LFDR) estimator for germline SNV calling. Our approach learns model parameters without prior information, and simultaneously accounts for information across all sites in the genomic regions of interest. We also propose another LFDR-based algorithm that reliably prioritizes a given list of mutations called by any other variant-calling algorithm. We use a suite of gold-standard cell line data to compare our LFDR approach against a collection of widely used, state of the art programs. We find that our LFDR approach approximately matches or exceeds the performance of all of these programs, despite some very large differences among them. Furthermore, when prioritizing other algorithms’ calls by our LFDR score, we find that by manipulating the type I-type II tradeoff we can select subsets of variant calls with minimal loss of sensitivity but dramatic increases in precision.

254. Genomic diversity of Gram-negative Bacilli from Bloodstream Infections in Hospitalized Children

Abstract Background Gram-negative bloodstream infection (BSI) in hospitalized children is associated with significant morbidity and mortality. It is unclear whether bacterial clonality of isolates during BSI in a unique patient is associated with clinical characteristics such as immunocompromise. We hypothesized that genomic diversity of BSI isolates would be higher in immunocompromised patients due to mucosal barrier injury and therefore larger translocation bottleneck. Methods We prospectively enrolled patients with a central venous catheter who were hospitalized at Boston Children’s Hospital and had a positive blood culture that grew E. coli, Klebsiella spp., or Pseudomonas aeruginosa. An aliquot from the blood culture bottle was plated on agar, and at least 24 colonies were picked randomly and subjected to whole-genome sequencing (WGS). Patient-specific reference genomes were constructed using long-read sequencing of at least 1 isolate on the PacBio platform. All isolates were sequenced using Illumina (MiSeq or NextSeq). Single nucleotide variant (SNV) calls were made using the WideVariant pipeline (github). Clinical characteristics were obtained retrospectively from the electronic medical record. We defined immunocompromise as neutropenia (ANC< 500) on the day of BSI or prior solid organ or stem cell transplant. The proportion of BSI episodes with clonal blood isolates (defined as 0 SNV) was compared between immunocompromised and non-immunocompromised patients using Fisher’s exact test. Results We enrolled 34 patients (median age 2.33 years, IQR 0.8-3.8 years), 25 having BSI from Enterobacterales (11 E. coli, 14 Klebsiella spp.) and 9 from Pseudomonas aeruginosa. Immunocompromised patients had a significantly higher proportion of Enterobacterales BSI episodes with clonal isolates compared with non-immunocompromised patients [13/18 (72%) vs. 1/7 (14%) respectively, P=0.021). This was not seen with Pseudomonas BSI, which had clonal isolates in 3/5 (60%) vs. 2/4 (50%) respectively (P=1). Conclusion Contrary to our hypothesis, immunocompromised patients were more likely to have clonal Enterobacterales BSI isolates compared with non-immunocompromised patients, perhaps due to lower host pressures for bacterial population diversification in immunocompromised patients. Disclosures Arolyn Conwill, PhD, Day Zero Diagnostics: current employee|Day Zero Diagnostics: Stocks/Bonds

Genomic data resources of the Brain Somatic Mosaicism Network for neuropsychiatric diseases

Somatic mosaicism is defined as an occurrence of two or more populations of cells having genomic sequences differing at given loci in an individual who is derived from a single zygote. It is a characteristic of multicellular organisms that plays a crucial role in normal development and disease. To study the nature and extent of somatic mosaicism in autism spectrum disorder, bipolar disorder, focal cortical dysplasia, schizophrenia, and Tourette syndrome, a multi-institutional consortium called the Brain Somatic Mosaicism Network (BSMN) was formed through the National Institute of Mental Health (NIMH). In addition to genomic data of affected and neurotypical brains, the BSMN also developed and validated a best practices somatic single nucleotide variant calling workflow through the analysis of reference brain tissue. These resources, which include >400 terabytes of data from 1087 subjects, are now available to the research community via the NIMH Data Archive (NDA) and are described here.

Early Identification of Refractory/Relapsed Diffuse Large B Cell Lymphoma with Serial Ctdna Sampling

Background: Diffuse Large B Cell Lymphoma (DLBCL) is an aggressive lymphoma that is curable in 60% of patients with chemoimmunotherapy. Outcomes are poor for those experiencing relapsed or refractory disease (rrDLBCL), particularly for patients with primary refractory disease (REFR, < 9 months from diagnosis) and early relapse (ER, 9 months to 2 years from diagnosis). Some of these patients would be candidates for chimeric antigen receptor T cell therapy (CART) in second line, where outcomes are superior when the disease burden is low. Therefore, exploring strategies to identify these high-risk patients early is important. Plasma circulating tumor DNA has been shown to be prognostic in various DLBCL cohorts. Method: We developed a custom panel of 170 genes to identify early treatment failure in a cohort of 171 patients that had profiling performed on 323 plasma samples. Plasma samples were taken serially starting at diagnosis and as patients progressed through frontline treatment. All plasma samples underwent DNA extraction, library preparation and subsequent sequencing using a panel of DLBCL related genes at high read depth (~1000x). Single nucleotide variant (SNV) calling was carried out using a custom pipeline including paired normal DNA for improved somatic variant detection. ctDNA fraction was estimated based on the highest variant allele fraction detected, using a loss of heterozygosity somatic model. Our ctDNA analysis focused on samples at diagnosis, cycle 2 of therapy, and end of treatment in order to identify early determinants of refractory disease. Changes in ctDNA levels between time points were represented as log2 ratio of ctDNA fraction. Results: rrDLBCL cases were separated into 3 categories based on the time between diagnosis and progressive disease (PD): 47 REFR, 34 ER, and 31 late relapse (LR, >24 months). The remaining patients were disease free after frontline therapy for over 24 months (CR, n=59), resulting in a cohort enriched for rrDLBCL (65% of cases). The average international prognostic index (IPI) of each group at diagnosis was REFR=3.38, ER=2.87, LR=2.81, and CR=2.41. Cell of Origin, as determined by Hans algorithm, showed a higher number of non-GCB samples in ER and LR groups (53% and 60%, respectively) while REFR and CR displayed increased GCB cases (63% and 66%, respectively). The REFR patients were significantly more likely to be of a 4/5 IPI score at diagnosis than other groups (p=0.0203). Progression-free survival (PFS2) at relapse therapy for REFR (median=0.29 years) was shorter when compared to both ER (p=0.0104, median=0.44 years) and LR (p=0.0012, median=0.80 years). Using the diagnostic plasma sample, there was no significant difference in ctDNA fraction in any of the three groups. When the sample with the highest ctDNA fraction was compared between patients, REFR had significantly higher levels than LR and CR, consistent with a higher overall tumor burden (p=0.001 and 0.026, respectively). There was a trend towards higher ctDNA fraction at the end of treatment in both REFR and ER groups. As ctDNA dynamics are known to be informative of molecular response, we compared patients using the change in ctDNA fraction at cycle 2 of therapy (log2 ratio). This value was significantly different in REFR patients (p=0.0073 vs ER), consistent with a lower rate of molecular response to treatment. Conclusions: While refractory and later relapsed DLBCL have similar ctDNA features at diagnosis, we find they have distinct ctDNA dynamics during treatment. REFR had higher maximal ctDNA levels across time points and both REFR and ER patients exhibited higher levels of ctDNA at end of treatment. Moreover, patients with minimal change in ctDNA levels at cycle 2 of therapy are at a high risk of treatment failure and refractory disease. Further exploration of the specific mutational patterns is ongoing. These results could enable the earlier identification of rrDLBCL and facilitate the prioritization of approaches for therapeutic intervention in rrDLBCL.

A comparative investigation of single nucleotide variant calling for a personal non-Caucasian sequencing sample.

Dropping cost and increasing clinical application of whole genome sequencing (WGS) lead a necessity of efficient (accurate and rapid) variant calling procedures from a personal WGS data (n = 1). A number of variant calling pipelines have been introduced utilizing the human genome reference GRCh38 as a reference and a benchmark dataset called 'NA12878', which are both 'standard' but limited ethnic origin. Considering the nature of variant calling algorithms and recent updates in sequencing protocol, however, it is necessary to revisit the efficiency of the current best pipelines for a personal WGS data from diverse ethnicity. We discuss the most efficient practices for variant calling of a personal WGS reads, with a particular emphasis on whether (1) ethnic match or mismatch between the reference genome and a WGS data produces a distinct result and more importantly (2) there is an ethnic-specific optimal workflow. Here, we generate an appropriate WGS data, DNA array, and sufficient number of Sanger validated variants from a single Korean subject to perform such a comprehensive comparison. We applied this WGS reads and the 'NA12878' reads to 8 different variant calling pipelines with 2 different reference genomes (GRCh38 and KOREF, a Korean reference genome) to which the WGS reads from different ethnic origins are aligned. We evaluated the performance of the pipelines with the matched array genotype data and Sanger sequencing validation and demonstrated that: regardless to the ethnic match/mismatch (1) Novoalign-GATK4 showed the most efficient performance with the exceptional calls in MHC region; (2) the overall performance was better with GRCh38, while a significant difference in recall was observed. In addition, we found it is largely reduced computing cost maintaining performance to remove 'markduplication' step with PCR-free WGS data. For variant calling of a personal PCR-free WGS data, regardless of ethnicity consideration, we recommend the use of the Novoalign + GATK4 with GRCh38 and without 'markduplication'.

Single-nucleotide variant calling in single-cell sequencing data with Monopogen.

Single-cell omics technologies enable molecular characterization of diverse cell types and states, but how the resulting transcriptional and epigenetic profiles depend on the cell's genetic background remains understudied. We describe Monopogen, a computational tool to detect single-nucleotide variants (SNVs) from single-cell sequencing data. Monopogen leverages linkage disequilibrium from external reference panels to identify germline SNVs and detects putative somatic SNVs using allele cosegregating patterns at the cell population level. It can identify 100 K to 3 M germline SNVs achieving a genotyping accuracy of 95%, together with hundreds of putative somatic SNVs. Monopogen-derived genotypes enable global and local ancestry inference and identification of admixed samples. It identifies variants associated with cardiomyocyte metabolic levels and epigenomic programs. It also improves putative somatic SNV detection that enables clonal lineage tracing in primary human clonal hematopoiesis. Monopogen brings together population genetics, cell lineage tracing and single-cell omics to uncover genetic determinants of cellular processes.

XAtlas: scalable small variant calling across heterogeneous next-generation sequencing experiments.

The growing volume and heterogeneity of next-generation sequencing (NGS) data complicate the further optimization of identifying DNA variation, especially considering that curated high-confidence variant call sets frequently used to validate these methods are generally developed from the analysis of comparatively small and homogeneous sample sets. We have developed xAtlas, a single-sample variant caller for single-nucleotide variants (SNVs) and small insertions and deletions (indels) in NGS data. xAtlas features rapid runtimes, support for CRAM and gVCF file formats, and retraining capabilities. xAtlas reports SNVs with 99.11% recall and 98.43% precision across a reference HG002 sample at 60× whole-genome coverage in less than 2 CPU hours. Applying xAtlas to 3,202 samples at 30× whole-genome coverage from the 1000 Genomes Project achieves an average runtime of 1.7 hours per sample and a clear separation of the individual populations in principal component analysis across called SNVs. xAtlas is a fast, lightweight, and accurate SNV and small indel calling method. Source code for xAtlas is available under a BSD 3-clause license at https://github.com/jfarek/xatlas.

Genotyping Microbial Communities with MIDAS2: From Metagenomic Reads to Allele Tables.

The Metagenomic Intra-Species Diversity Analysis System 2 (MIDAS2) is a scalable pipeline that identifies single nucleotide variants and gene copy number variants in metagenomes using comprehensive reference databases built from public microbial genome collections (metagenotyping). MIDAS2 is the first metagenotyping tool with functionality to control metagenomic read mapping filters and to customize the reference database to the microbial community, features that improve the precision and recall of detected variants. In this article we present four basic protocols for the most common use cases of MIDAS2, along with supporting protocols for installation and use. In addition, we provide in-depth guidance on adjusting command line parameters, editing the reference database, optimizing hardware utilization, and understanding the metagenotyping results. All the steps of metagenotyping, from raw sequencing reads to population genetic analysis, are demonstrated with example data in two downloadable sequencing libraries of single-end metagenomic reads representing a mixture of multiple bacterial species. This set of protocols empowers users to accurately genotype hundreds of species in thousands of samples, providing rich genetic data for studying the evolution and strain-level ecology of microbial communities. © 2022 The Authors. Current Protocols published by Wiley Periodicals LLC. Basic Protocol 1: Species prescreening Basic Protocol 2: Download MIDAS reference database Basic Protocol 3: Population single nucleotide variant calling Basic Protocol 4: Pan-genome copy number variant calling Support Protocol 1: Installing MIDAS2 Support Protocol 2: Command line inputs Support Protocol 3: Metagenotyping with a custom collection of genomes Support Protocol 4: Metagenotyping with advanced parameters.

Molecular Characterization of Clinical Response in Relapsed/Refractory Acute Myeloid Leukemia and High-Risk Myelodysplastic Syndrome Patients Treated with Single Agent Emavusertib

Introduction Acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) are myeloid malignancies that exhibit a dynamic mutational landscape as the disease progresses. Genetic mutations in the splicing factors SF3B1 and U2AF1 drive overexpression of a highly active long isoform of interleukin-1 receptor-associated kinase 4 (IRAK4), which is critical in triggering inflammation, oncogenesis, and survival of cancer cells. Splicing factor mutations are frequently reported in AML and MDS and have been associated with disease progression and poor prognosis in relapsed/refractory (R/R) diseases. Emavusertib is a potent oral inhibitor of IRAK4 as well as Fms-like tyrosine kinase 3 (FLT3). The safety, clinical activity, and potential biomarkers of emavusertib in AML or high-risk MDS (HR-MDS) are investigated in the ongoing open-label, phase 1/2a TakeAim Leukemia trial (NCT04278768). Methods Eligible patients with R/R AML or HR-MDS were enrolled and treated with emavusertib. The exploratory objectives of the study are to assess the association between clinical activity and target-related biomarkers, gene expression signatures, and molecular subtypes. Bone marrow or peripheral blood were collected at the baseline and on treatment. To characterize the mutational profiles of enrolled patients and further define the targeted population, targeted next generation sequencing (NGS) of 68 genes was performed on genomic DNA from bone marrow or peripheral blood mononuclear cells. A minimum coverage of 100x, variant allele frequency (VAF) of >1%, and a VARSCAN-assigned variant p<0.001 were used as thresholds for single nucleotide variants (SNV) calling and presumed somatic driver mutations were identified as previously reported (Metzeler et al. Blood 2016). Mutations from patients’ molecular pathology reports were also documented. Results As of December 16, 2021, 49 patients have been treated with emavusertib monotherapy at one of dose levels of 200, 300, 400, or 500 mg BID, including 30 R/R AML and 19 R/R HR-MDS. The median number of prior anti-cancer therapies was 2 (range 1-5). This abstract is intended to add additional molecular findings to the patient subset presented in 2022 ASCO, of which 6 AML and 7 HR-MDS had splicing factor SF3B1 or U2AF1 mutations and 3 AML had FLT3 mutations, including 2 AML with both splicing factor and FLT3 mutations (hereafter referred to as "targeted mutations"). In the patients with targeted mutations, with available mutational profile data or molecular reports from sites, there were 6 responders: 1 CR, 1 CRh, and 4 mCR. In this responder population, the most common co-mutation was ASXL1. During emavusertib treatment, FLT3-mutated clones became undetectable in 2 of 3 (67%) FLT3-mutated AML patients. One of the 3 patients (33%) with a FLT3 mutation achieved CR and the remaining 2 patients achieved SD (1 with > 90% blast reduction). Additional gene mutations, such as ASXL1, RUNX1, CEBPA, ETNK1, ETV6, and SRSF2, became undetectable during emavusertib treatment. New mutations that emerged after ≥ 1 cycle of emavusertib treatment include NRAS, SETBP1, RUNX1, WT1, STAG2, KRAS, CCND2, DNMT3A, ASXL1, and NOTCH1. Analysis of genomic and transcriptomic profiles will be updated. Conclusion Emavusertib demonstrated anti-cancer activity in heavily pretreated AML and HR-MDS patients, especially in those with U2AF1, SF3B1, or FLT3 mutations. Our preliminary NGS data is suggestive of molecular responses and disease-modifying activity of emavusertib. Patients with targeted mutations responded to emavusertib, even in the presence of co-mutations of ASXL1 that are typically associated with poor prognosis. We will continue to explore potential genomic biomarkers of emavusertib sensitivity and resistance

Genomic Landscaping of Post-Transplant Lymphoproliferative Disorders Using Circulating Tumor DNA

Introduction: Identification of post-transplant lymphoproliferative disorder (PTLD) is a diagnostic challenge. The current diagnostic framework usually require analysis on tissue and imaging through PET/CT scans. Yet, these methods have diagnostic limitations. For instance due to missing information on tumor heterogeneity and relatively low sensitivity of PET/CT scans. Measuring cell free DNA (cfDNA) represents a novel, minimally-invasive method for tumor detection and genotyping. In this study we explored the feasibility of plasma cfDNA tumor genotyping with targeted deep sequencing to detect copy number variation (CNV), single nucleotide variants (SNVs) and Epstein Bar Virus (EBV) sequences, using a new in-house variant calling pipeline. Methods: In this multi-center observational cohort study of 18 patients diagnosed with monomorphic PTLD, we performed deep targeted NGS for SNV detection and low-coverage whole genome sequencing (lcWGS) for CNV detection. NGS was targeted on a panel of genes (n = 39) known to be recurrently mutated in B-cell lymphomas (244 kbp). All patients gave written informed consent. The R package CNAclinic was used to identify genome wide CNAs and fraction of the genome altered (FGA). The ichorCNA package was used to determine the ctDNA fraction in cfDNA samples. SNVs were called using a pipeline, combining 4 (cfDNA) SNV calling tools: Mutect2, LoFreq, VarDict & SiNVICT. Variants were filtered on presumed functional effect (Synonymous-, UTR-, Intron-, 2 kb upstream-, 2 kb downstream variants were removed). A "panel of normals” (variant blacklist) was generated from variants found in cfDNA from healthy controls, which was used to produce a variant blacklist to filter out technical artifacts and SNPs. Finally, variants with a variant allele frequency (VAF) < 1% and variants with an allele frequency (AF) of > 1% in 1000Genomes and/or gnomAD were filtered out. Results: The cohort included 18 PTLD patients. The majority of patients had stage IV disease (n=13, 76%). All of the PTLD cases were histologically proven, of diffuse large B cell lymphoma type and 10/18 patients (55%) were considered to have an EBV-negative tumor. EBV copy number in plasma was performed by qPCR in 12 patients and was considered to be elevated in four patients (>5000 copies/mL). The fraction of reads mapping to the EBV genome was compared between controls (n=8), EBV-negative PTLD (n=10) and EBV-positive PTLD (n=8). The median fraction of EBV reads for controls and EBV-negative PTLD were similar with a fraction of 1.5x10-4 (Q1-Q3: 1.3x10-4-1.6x10-4) and 1.8x10-4 (Q1-Q3: 1.4x10-4-3.9x10-4), respectively (p>0.05, Kruskal-Wallis). In EBV-positive tumors, the median fraction of EBV reads was 3-log higher, 5.3x10-1 (Q1-Q3: 7.3 x10-2-2) (p≤0.05, Kruskal-Wallis) CNVs were detected in cfDNA of 11/18 (61%) patients. Two of the patients in whom no CNVs were detected had initiated therapy at the time of plasma sampling. Median fraction of genome altered was higher in EBV-negative tumors (0.03, Q1-Q3: 2.5x10-4-0.12) than in EBV-positive tumors (5x10-4, Q1-Q3: 0-0.03), but this difference was not significant (p>0.05, Mann-Whitney U) (Figure 1). In the top 15 recurrently mutated genes (described below), across all 18 samples, 180 SNVs were found in total. The median amount of SNVs found in each sample was 11 (range 6-25). The most frequently mutated gene was CREPPB (12/18, 67%), in which the most prevalent mutation was p.K1497T (8/18, 44%). ARID1A was mutated in 10/18 cases (56%), followed by KMT2D (9/18, 50%). TET2, TP53, MEF2B & TNFAIP3 were all mutated in 7/18 cases (39%). SPEN & IRF2BP2 were mutated in 6/18 cases (33%). PIM1, NOTCH1, ATM, CARD11, DDX3X, IGLL5 were all mutated in 5/18 cases (28%) (Figure 2). Few of the mutations that are called in the top 15 recurrently mutated genes (8/180, 0.04%) lay close to 50% VAF, which might imply that these are SNPs not filtered out by the pipeline. Our SNV findings are comparable with previous literature, in which KMT2D, TET2 & TP53 are also amongst top-mutated genes (Menter et al., 2017), highlighting disruption of epigenetic regulation, NFkB pathway, and DNA damage response as drivers of PTLD. Conclusion: lcWGS in combination deep targeted sequencing allows for detection of CNV an SNV in diagnostic samples of patients with monomorphic PTLD. These data can potentially be used at screening, diagnosis and response monitoring, which is part of the ongoing prospective study NTR 7402. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Favored single nucleotide variants identified using whole genome Re-sequencing of Austrian and Chinese cattle breeds

Cattle have been essential for the development of human civilization since their first domestication few thousand years ago. Since then, they have spread across vast geographic areas following human activities. Throughout generations, the cattle genome has been shaped with detectable signals induced by various evolutionary processes, such as natural and human selection processes and demographic events. Identifying such signals, called selection signatures, is one of the primary goals of population genetics. Previous studies used various selection signature methods and normalized the outputs score using specific windows, in kbp or based on the number of SNPs, to identify the candidate regions. The recent method of iSAFE claimed for high accuracy in pinpointing the candidate SNPs. In this study, we analyzed whole-genome resequencing (WGS) data of ten individuals from Austrian Fleckvieh (Bos taurus) and fifty individuals from 14 Chinese indigenous breeds (Bos taurus, Bos taurus indicus, and admixed). Individual WGS reads were aligned to the cattle reference genome of ARS. UCD1.2 and subsequently undergone single nucleotide variants (SNVs) calling pipeline using GATK. Using these SNVs, we examined the population structure using principal component and admixture analysis. Then we refined selection signature candidates using the iSAFE program and compared it with the classical iHS approach. Additionally, we run Fst population differentiation from these two cattle groups. We found gradual changes of taurine in north China to admixed and indicine to the south. Based on the population structure and the number of individuals, we grouped samples to Fleckvieh, three Chinese taurines (Kazakh, Mongolian, Yanbian), admixed individuals (CHBI_Med), indicine individuals (CHBI_Low), and a combination of admixed and indicine (CHBI) for performing iSAFE and iHS tests. There were more significant SNVs identified using iSAFE than the iHS for the candidate of positive selection and more detectable signals in taurine than in indicine individuals. However, combining admixed and indicine individuals decreased the iSAFE signals. From both within-population tests, significant SNVs are linked to the olfactory receptors, production, reproduction, and temperament traits in taurine cattle, while heat and parasites tolerance in the admixed individuals. Fst test suggests similar patterns of population differentiation between Fleckvieh and three Chinese taurine breeds against CHBI. Nevertheless, there are genes shared only among the Chinese taurine, such as PAX5, affecting coat color, which might drive the differences between these yellowish coated breeds, and those in the greater Far East region.

Identification of a signature of evolutionarily conserved stress-induced mutagenesis in cancer.

The clustering of mutations observed in cancer cells is reminiscent of the stress-induced mutagenesis (SIM) response in bacteria. Bacteria deploy SIM when faced with DNA double-strand breaks in the presence of conditions that elicit an SOS response. SIM employs DinB, the evolutionary precursor to human trans-lesion synthesis (TLS) error-prone polymerases, and results in mutations concentrated around DNA double-strand breaks with an abundance that decays with distance. We performed a quantitative study on single nucleotide variant calls for whole-genome sequencing data from 1950 tumors, non-inherited mutations from 129 normal samples, and acquired mutations in 3 cell line models of stress-induced adaptive mutation. We introduce statistical methods to identify mutational clusters, quantify their shapes and tease out the potential mechanism that produced them. Our results show that mutations in both normal and cancer samples are indeed clustered and have shapes indicative of SIM. Clusters in normal samples occur more often in the same genomic location across samples than in cancer suggesting loss of regulation over the mutational process during carcinogenesis. Additionally, the signatures of TLS contribute the most to mutational cluster formation in both patient samples as well as experimental models of SIM. Furthermore, a measure of cluster shape heterogeneity was associated with cancer patient survival with a hazard ratio of 5.744 (Cox Proportional Hazard Regression, 95% CI: 1.824–18.09). Our results support the conclusion that the ancient and evolutionary-conserved adaptive mutation response found in bacteria is a source of genomic instability in cancer. Biological adaptation through SIM might explain the ability of tumors to evolve in the face of strong selective pressures such as treatment and suggests that the conventional ‘hit it hard’ approaches to therapy could prove themselves counterproductive.

P606: META-ANALYSIS OF LOW ALLELIC BURDEN C481 BTK-MUTATIONS CALLS FOR CAUTION IN IBRUTINIB RESISTANCE EVALUATION

Background: Introduction of the orally administered tyrosine kinase inhibitor ibrutinib, a drug approved for relapsed or refractory mantle cell lymphoma in late 2013 and chronic lymphocytic leukemia in 2014, has changed treatment algoritms for B-cell malignancies and improved progression-free survival. Ibrutinib irreversibly targets Bruton’s tyrosine kinase (BTK), thus interrupting B-cell receptor signaling. However, it is known that mutations in the BTK gene, located on the X chromosome, can mediate acquired resistance, specifically modifying cysteine residue 481 to serine (C481S, X:101356176, GRCh38). Aims: Despite this knowledge, it is still unknown how low mutational burden can contribute to general resistance towards ibrutinib treatment. Thus, we performed a meta-analysis based on previous cardinal studies to investigate the collective and gender-specific distribution of variant allele frequencies. Methods: We implemented meta-analysis of single-nucleotide variant calls from seven studies involving whole-exome sequencing (Woyach et al. 2014 and 2017, Maddocks et al. 2015, Ahn et al. 2017, Kanagal-Shamanna et al. 2019, Gángó et al. 2020, Bödör et al. 2021) on mutational profiling concerning ibrutinib resistance and CLL. Four of these studies contained gender specific data. Computational analyses, statistics, and visualization were performed in Wolfram Language, R, and GraphPad Prism. Results: From the available data included in these studies, we were able to base our analyses on 226 non-BTK mutations and 170 BTK mutations within a variant allele frequency (VAF) range of 0.0001–0.9, of which 93 mutation candidates from males (VAF < 0.9) and 77 from females (VAF < 0.44) were observed. Extrapolative analysis of allele frequency distribution from all other variants (excl. BTK) in the studies revealed an exponential increase in the number of low burden variants below 10% VAF (R2>0.99). These variants were not significantly different between genders (PMann-Whitney=0.24, nmale=130, nfemale=73). In contrast, a significant difference was observed between male variant allele frequencies (PMann-Whitney=0.005) and females owing to BTK hemizygosity in males. However, this significance persisted when female BTK mutations were transformed to reflect a single copy (VAF/2, PMann-Whitney=0.015, ΔVAF=0.05) for gender comparison and even when observations for males were alternatively transformed to reflect two copies (2*VAF, PMann-Whitney<0.001, ΔVAF=0.19). The previously discovered threshold of 10% was set as the detection limit. Summary/Conclusion: Defining specific molecular contributions in drug resistance is imperative for the correct use of targeted therapy. It has been questioned how the low allelic burden of BTK mutations plays a role. Therefore, it is crucial to investigate the lower limit of detection. The meta-analysis suggests that variants below a certain threshold, here empirically 10%, may in many cases be false-positive variants, in agreement with other reports involving WES in general, introduced by erroneous base calling, PCR errors, etc. Although intriguing, it remains to elaborate whether the haploinsufficiency of BTK in males skew in the degree of ibrutinib resistance.

Combinatorial and Machine Learning Approaches for Improved Somatic Variant Calling From Formalin-Fixed Paraffin-Embedded Genome Sequence Data.

Formalin fixation of paraffin-embedded tissue samples is a well-established method for preserving tissue and is routinely used in clinical settings. Although formalin-fixed, paraffin-embedded (FFPE) tissues are deemed crucial for research and clinical applications, the fixation process results in molecular damage to nucleic acids, thus confounding their use in genome sequence analysis. Methods to improve genomic data quality from FFPE tissues have emerged, but there remains significant room for improvement. Here, we use whole-genome sequencing (WGS) data from matched Fresh Frozen (FF) and FFPE tissue samples to optimize a sensitive and precise FFPE single nucleotide variant (SNV) calling approach. We present methods to reduce the prevalence of false-positive SNVs by applying combinatorial techniques to five publicly available variant callers. We also introduce FFPolish, a novel variant classification method that efficiently classifies FFPE-specific false-positive variants. Our combinatorial and statistical techniques improve precision and F1 scores compared to the results of publicly available tools when tested individually.

Accurate and scalable variant calling from single cell DNA sequencing data with ProSolo

Accurate single cell mutational profiles can reveal genomic cell-to-cell heterogeneity. However, sequencing libraries suitable for genotyping require whole genome amplification, which introduces allelic bias and copy errors. The resulting data violates assumptions of variant callers developed for bulk sequencing. Thus, only dedicated models accounting for amplification bias and errors can provide accurate calls. We present ProSolo for calling single nucleotide variants from multiple displacement amplified (MDA) single cell DNA sequencing data. ProSolo probabilistically models a single cell jointly with a bulk sequencing sample and integrates all relevant MDA biases in a site-specific and scalable—because computationally efficient—manner. This achieves a higher accuracy in calling and genotyping single nucleotide variants in single cells in comparison to state-of-the-art tools and supports imputation of insufficiently covered genotypes, when downstream tools cannot handle missing data. Moreover, ProSolo implements the first approach to control the false discovery rate reliably and flexibly. ProSolo is implemented in an extendable framework, with code and usage at: https://github.com/prosolo/prosolo

A benchmarking study of SARS-CoV-2 whole-genome sequencing protocols using COVID-19 patient samples

SummarySevere acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is an emerging new type of coronavirus that is responsible for the COVID-19 pandemic and the unprecedented global health emergency. Whole-genome sequencing (WGS) of SARS-CoV-2 plays a critical role in understanding the disease. Performance variation exists across SARS-CoV-2 viral WGS technologies, but there is currently no benchmarking study comparing different WGS sequencing protocols. We compared seven different SARS-CoV-2 WGS library protocols using RNA from patient nasopharyngeal swab samples under two storage conditions with low and high viral inputs. We found large differences in mappability and genome coverage, and variations in sensitivity, reproducibility, and precision of single-nucleotide variant calling across different protocols. For certain amplicon-based protocols, an appropriate primer trimming step is critical for accurate single-nucleotide variant calling. We ranked the performance of protocols based on six different metrics. Our findings offer guidance in choosing appropriate WGS protocols to characterize SARS-CoV-2 and its evolution.

CNVfilteR: an R/Bioconductor package to identify false positives produced by germline NGS CNV detection tools.

SummaryGermline copy-number variants (CNVs) are relevant mutations for multiple genetics fields, such as the study of hereditary diseases. However, available benchmarks show that all next-generation sequencing (NGS) CNV calling tools produce false positives. We developed CNVfilteR, an R package that uses the single-nucleotide variant calls usually obtained in germline NGS pipelines to identify those false positives. The package can detect both false deletions and false duplications. We evaluated CNVfilteR performance on callsets generated by 13 CNV calling tools on three whole-genome sequencing and 541 panel samples, showing a decrease of up to 44.8% in false positives and consistent F1-score increase. Using CNVfilteR to detect false-positive calls can improve the overall performance of existing CNV calling pipelines.Availability and implementationCNVfilteR is released under Artistic-2.0 License. Source code and documentation are freely available at Bioconductor (http://www.bioconductor.org/packages/CNVfilteR).Supplementary information Supplementary data are available at Bioinformatics online.