WES Datasets Research Articles

Mugen-UMAP: UMAP visualization and clustering of mutated genes in single-cell DNA sequencing data

BackgroundThe application of Uniform Manifold Approximation and Projection (UMAP) for dimensionality reduction and visualization has revolutionized the analysis of single-cell RNA expression and population genetics. However, its potential in single-cell DNA sequencing data analysis, particularly for visualizing gene mutation information, has not been fully explored.ResultsWe introduce Mugen-UMAP, a novel Python-based program that extends UMAP’s utility to single-cell DNA sequencing data. This innovative tool provides a comprehensive pipeline for processing gene annotation files of single-cell somatic single-nucleotide variants and metadata to the visualization of UMAP projections for identifying clusters, along with various statistical analyses. Employing Mugen-UMAP, we analyzed whole-exome sequencing data from 365 single-cell samples across 12 non-small cell lung cancer (NSCLC) patients, revealing distinct clusters associated with histological subtypes of NSCLC. Moreover, to demonstrate the general utility of Mugen-UMAP, we applied the program to 9 additional single-cell WES datasets from various cancer types, uncovering interesting patterns of cell clusters that warrant further investigation. In summary, Mugen-UMAP provides a quick and effective visualization method to uncover cell cluster patterns based on the gene mutation information from single-cell DNA sequencing data.ConclusionsThe application of Mugen-UMAP demonstrates its capacity to provide valuable insights into the visualization and interpretation of single-cell DNA sequencing data. Mugen-UMAP can be found at https://github.com/tengchn/Mugen-UMAP

Development and validation of a reliable DNA copy-number-based machine learning algorithm (CopyClust) for breast cancer integrative cluster classification

The Integrative Cluster subtypes (IntClusts) provide a framework for the classification of breast cancer tumors into 10 distinct groups based on copy number and gene expression, each with unique biological drivers of disease and clinical prognoses. Gene expression data is often lacking, and accurate classification of samples into IntClusts with copy number data alone is essential. Current classification methods achieve low accuracy when gene expression data are absent, warranting the development of new approaches to IntClust classification. Copy number data from 1980 breast cancer samples from METABRIC was used to train multiclass XGBoost machine learning algorithms (CopyClust). A piecewise constant fit was applied to the average copy number profile of each IntClust and unique breakpoints across the 10 profiles were identified and converted into ~ 500 genomic regions used as features for CopyClust. These models consisted of two approaches: a 10-class model with the final IntClust label predicted by a single multiclass model and a 6-class model with binary reclassification in which four pairs of IntClusts were combined for initial multiclass classification. Performance was validated on the TCGA dataset, with copy number data generated from both SNP arrays and WES platforms. CopyClust achieved 81% and 79% overall accuracy with the TCGA SNP and WES datasets, respectively, a nine-percentage point or greater improvement in overall IntClust subtype classification accuracy. CopyClust achieves a significant improvement over current methods in classification accuracy of IntClust subtypes for samples without available gene expression data and is an easily implementable algorithm for IntClust classification of breast cancer samples with copy number data.

Novel Loci for Alzheimer Disease Identified by Genome Wide Association Study in Ashkenazi Jews

AbstractBackgroundPrevious studies identified associations of Alzheimer’s disease (AD) risk primarily in outbred European ancestry (EA) populations. We focused on Ashkenazi Jews (AJs) who descended from a founder population in eastern Europe.MethodWe discriminated AJs in the multiethnic Alzheimer’s Disease Genetic Consortium (ADGC) TOPMed imputed GWAS dataset (n = 53,502) and whole genome sequencing (WGS, n = 16,815) and whole exome sequencing (WES, n = 20,504) datasets assembled by the Alzheimer’s Disease Sequencing Project using a reference GWAS dataset including persons with four AJ grandparents (n = 3,435). Principal component (PC) analysis was performed for each AD dataset combined with the AJ reference data, and a Gaussian mixture model was applied for population clustering. GWA analyses were performed by combining AJs discriminated from the ADGC and WGS subjects (G1) using genome‐wide data and from all three datasets using variants called from exome capture (G2). Separate analyses of individual common (MAF≥1%) and rare (MAF<1%) variants and gene‐based analyses including rare variants were performed using models adjusting for age, sex, and PCs.ResultTotals of 2,956, 738 and 1,062 unique AJs were identified in the ADGC, WGS and WES datasets, respectively. After filtering duplicate and closely related individuals, the sample sizes were 1,355 cases and 1,661 controls for G1 and 1,504 cases and 2,047 controls for G2 analyses. Excluding the APOE region, G1 analyses revealed a genome‐wide significant (GWS) association (P<5.0×10−8) with the TREM2 R47H missense mutation (P = 9.66×10−9) and suggestive associations with rs541586606 near RAB3B (P = 5.01×10−8), rs545690149 near ZNF890P, (P = 5.77×10−8) and rs1225737296 near ANXA10 (P = 6.32×10−8). In G2 analyses, we obtained GWS associations with GBA intronic SNP rs3115534 (P = 3.20×10−13) and TREM2 R47H (P = 2.64×10−10), and suggestive associations with rs1003710 in SMAP2 (P = 1.91×10−7) and rs200698976 in ZNF890P (P = 3.49×10−7). No GWS rare variant associations were identified. Significant gene‐based association was identified with GIPR (7.34×10−7).ConclusionAmong the novel findings, GBA was previously associated with Parkinson disease and Lewy body dementia (LBD). Remarkably, rs3115534 is located 35 bp from, but not in linkage disequilibrium with, N370S that is associated with LBD in AJs. Moreover, rs3115534 (MAF = 0.027 AJ controls) is rare in EAs (MAF = 0.0015). Our results highlight the efficacy of GWAS in founder populations.

Multiple Viruses Detected in Human DNA are Associated with Alzheimer Disease Risk

AbstractBackgroundMultiple infectious agents, including viruses, bacteria, fungi, and protozoa, have been linked to Alzheimer disease (AD) by independent lines of evidence.MethodWe mapped and quantified DNA sequence reads from whole exome and whole genome sequencing data in 37,000 AD cases and controls from of the Alzheimer Disease Sequencing Project that did not align to the human genome to 318 viral reference sequences. After quality control, the sample included 1,696 samples derived from brain (cases=1,314, controls=382) and 14,588 derived from blood (cases=5,813, controls=8,775). We tested individual and cumulative normalized viral read counts, both free and inserted in the host genome, for association with AD using multiple machine learning (ML) classifiers, selecting the top 27 viral species according to an algorithm based on the species’ predictive rank weighted by the model’s overall predictive accuracy across models. These top 27 viruses were tested for association with AD using logistic regression models adjusted for PCR amplification, sequencing site, sex, APOE genotype, and ancestry. Analyses were stratified by tissue source.ResultWe detected sequence reads from 175 unique viral species, herpes simplex virus 1 (HSV‐1) being the most common. Viral read counts varied significantly by sequence center, tissue source, sex, PCR use, and ancestry in both the WES and WGS datasets. ML results indicated that herpes viruses (1, 2, 4, 5, 6A, 6B, 7, and 8), HIV, Hepatitis C, Molluscum contagiosum, several Torque Teno virus species (TTV), HPV‐71, Coronavirus 229E, Tick‐borne encephalitis, human Mastedenovirus, human polyomavirus 2, variola virus, primate t‐lymphotropic virus 1, human endogenous retrovirus K, and total normalized mapped viral reads were significant predictors of AD. Subsequent meta‐analyses of results stratified by tissue source showed that AD was significantly associated with HSV‐1(OR=4.07, P=2.22x10‐5), Human papillomavirus type 71(OR=3.26, P=0.006), TTV‐10 (OR=22.11, P=0.004) , and cumulative viral read count (OR=7.34, P=0.03). After multiple testing correction, HSV‐1 remains significant.ConclusionThese results support the hypothesis that viral infections, especially HSV‐1, is associated with AD risk and demonstrate the potential of using deep sequencing technology to detect microbial agents in multiple tissues and discerning here‐to‐for unrecognized associations of infectious agents and AD.

Systematic benchmark of state-of-the-art variant calling pipelines identifies major factors affecting accuracy of coding sequence variant discovery

BackgroundAccurate variant detection in the coding regions of the human genome is a key requirement for molecular diagnostics of Mendelian disorders. Efficiency of variant discovery from next-generation sequencing (NGS) data depends on multiple factors, including reproducible coverage biases of NGS methods and the performance of read alignment and variant calling software. Although variant caller benchmarks are published constantly, no previous publications have leveraged the full extent of available gold standard whole-genome (WGS) and whole-exome (WES) sequencing datasets.ResultsIn this work, we systematically evaluated the performance of 4 popular short read aligners (Bowtie2, BWA, Isaac, and Novoalign) and 9 novel and well-established variant calling and filtering methods (Clair3, DeepVariant, Octopus, GATK, FreeBayes, and Strelka2) using a set of 14 “gold standard” WES and WGS datasets available from Genome In A Bottle (GIAB) consortium. Additionally, we have indirectly evaluated each pipeline’s performance using a set of 6 non-GIAB samples of African and Russian ethnicity. In our benchmark, Bowtie2 performed significantly worse than other aligners, suggesting it should not be used for medical variant calling. When other aligners were considered, the accuracy of variant discovery mostly depended on the variant caller and not the read aligner. Among the tested variant callers, DeepVariant consistently showed the best performance and the highest robustness. Other actively developed tools, such as Clair3, Octopus, and Strelka2, also performed well, although their efficiency had greater dependence on the quality and type of the input data. We have also compared the consistency of variant calls in GIAB and non-GIAB samples. With few important caveats, best-performing tools have shown little evidence of overfitting.ConclusionsThe results show surprisingly large differences in the performance of cutting-edge tools even in high confidence regions of the coding genome. This highlights the importance of regular benchmarking of quickly evolving tools and pipelines. We also discuss the need for a more diverse set of gold standard genomes that would include samples of African, Hispanic, or mixed ancestry. Additionally, there is also a need for better variant caller assessment in the repetitive regions of the coding genome.

Multiple viruses detected in human DNA are associated with Alzheimer disease risk.

Multiple infectious agents, including viruses, bacteria, fungi, and protozoa, have been linked to Alzheimer disease (AD) by independent lines of evidence. We quantified DNA sequence reads that align to viral and bacterial reference sequences present in whole exome (WES, N= 15,125) and whole genome (WGS, N= 4,443) sequences measured in blood (N=17,817) and brain (N=1,751) tissue from non-Hispanic white, African American and Caribbean Hispanic participants of the AD Sequencing Project including 9,380 AD cases and 10,188 controls. We tested the association of individual and cumulative normalized viral read counts with AD using logistic and LASSO regression models adjusting for tissue source, sequencing site, sex, APOE genotype, and population. Additional models were tested in subsets stratified by WES/WGS and tissue source. We used PathSeq, a software tool that performs subtraction on high-throughput sequencing data, to identify inserted microbial DNA and tested those species for association with AD. We detected sequences from 175 unique species of human viruses and 17,219 bacteria. Many individual and the cumulative viral read counts varied significantly by sequence center, tissue source, sex, and ancestry. Cumulative normalized human viral read count was associated with AD in every subset of the data (ORmeta = 1.05 per 100,000 units, Pmeta < 7.81x10-12 ). In the WES dataset, the presence of inserted herpes virus 6A sequence measured by PathSeq was protective against AD (OR=0.41, P=0.0007). We also found significant associations using LASSO. In the WES dataset, Hepatitis C was associated with increased AD risk. Herpes 6B was associated with decreased AD risk in both the WES and WGS datasets. In brain samples, herpes 4 and 6B were negatively associated with AD in both WES and WGS datasets, whereas herpes 8 was positively associated with AD and human herpes 3 was negatively associated in the WES dataset. No bacterial species were significantly associated with AD. These results are consistent with the hypothesis that viral infections contribute to AD pathogenesis and highlight the utility of next generation sequencing to detect microbial agents in multiple tissues. Contrary to previous results, except for human herpes 8, herpes viral DNA species are protective against AD.

Abstract 618: Spatio-temporal multiomics analysis reveals distinct molecular features in recurrent stage I non-small cell lung cancers after R0 tumor resection

Abstract A portion of stage I non-small cell lung cancer (NSCLC) patients who have received R0 resection experience unexpected tumor recurrence during follow up. Some of them even recurred within 12 months after surgery, which is unable to predict by any clinical characteristics. Few study investigates the molecular features of such short-term recurrent tumors. Multiomics analysis were performed on 81 stage I NSCLC patients who had received R0 resection, based on deep whole-exome sequencing (average depth &amp;gt;500x) on multi-region 239 tumor tissue samples and 81 matched-adjacent normal specimens, as well as RNA sequencing and plasma targeted circulating tumor DNA detection. Dynamic dimension comparison of genomic alteration features between tumor recurrence and non-recurrence patients was made according to 60-months median time of follow up. Published WES datasets of Stage I NSCLCs of 131 East Asian (EAS LUADs), 62 European (TRACERx) and 277 American patients (TCGA) were extracted as validation. 60.5% patients were non-smokers. Lower chromosomal instability was found in our cohort characterized by less trunk CNVs and lower CNV coverage on genome as compared to the European cohort. Age-related signature mutations significantly accumulated in recurrence-free tumors. Somatic mutations occurred in DNA damage repair pathways were more likely to cause tumor recurrence (p=0.04), particularly in homologous recombination (HRR) (p=0.017) and translesion synthesis (TLS) (p=0.009) pathways. Dynamic analysis discovered the essential genes in Fanconi anemia (FA), TLS and HRR pathways that coordinate together for DNA interstrand crosslink (ICL) recognition and faithful repair of DNA double strand breaks (DSBs) during DNA replication stress frequently mutated in early tumor recurrence patients (n=7/11, 63%). Twenty-two focal CNV-driver genes, including cell proliferation-related RECQL4 (8q24.3), PIK3CA (3q26.32), BCL9 (1q21.2), TBL1XR1 (3q26.32) frequently amplified in recurrent tumors, as well as BCL3 and CBLC (19q13.31-32) in the early recurrent tumors. RNA-seq data demonstrated cell cycle pathway significantly up-regulated in recurrent tumors. More TMB, CNV fraction and genome doubling events in FA-TLS-HRR alteration patients verified genomic instability caused by dysfunction of cell cycle control and ICL-DSB repair contributing to early tumor recurrence. Integrative model cross clinical and molecular features stratified stage I patient prognosis in both our cohort and European cohort, in which and high risk patients were more detected by circulating tumor DNA. This is the first study to dynamically investigate the genomic features specifically in stage I lung NSCLC according to the precise tumor recurrence time, which provides important clues to postoperative treatment strategy and drug target research. Citation Format: Kezhong Chen, Airong Yang, Shuangxiu Wu, Yuntao Nie, Haifeng Shen, Jian Bai, Lin Wu, Fan Yang, Jun Wang. Spatio-temporal multiomics analysis reveals distinct molecular features in recurrent stage I non-small cell lung cancers after R0 tumor resection [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 618.

WEScover: selection between clinical whole exome sequencing and gene panel testing

BackgroundWhole exome sequencing (WES) is widely adopted in clinical and research settings; however, one of the practical concerns is the potential false negatives due to incomplete breadth and depth of coverage for several exons in clinically implicated genes. In some cases, a targeted gene panel testing may be a dependable option to ascertain true negatives for genomic variants in known disease-associated genes. We developed a web-based tool to quickly gauge whether all genes of interest would be reliably covered by WES or whether targeted gene panel testing should be considered instead to minimize false negatives in candidate genes.ResultsWEScover is a novel web application that provides an intuitive user interface for discovering breadth and depth of coverage across population-scale WES datasets, searching either by phenotype, by targeted gene panel(s) or by gene(s). Moreover, the application shows metrics from the Genome Aggregation Database to provide gene-centric view on breadth of coverage.ConclusionsWEScover allows users to efficiently query genes and phenotypes for the coverage of associated exons by WES and recommends use of panel tests for the genes with potential incomplete coverage by WES.

Abstract 995: Integrative genomic characterization and CRISPR-mediated gene editing studies identify IRAKs as novel therapeutic targets for inflammation-driven prostate tumorigenesis

Abstract A major drawback to accurate diagnosis and treatment of inflammation-driven prostate cancer (PCa) progression is our limited understanding of the molecular mechanisms underlying aberrant inflammation signaling in PCa patients. Since asymptomatic chronic inflammation is hard to diagnose, we do not have adequate knowledge of how specific genes within the inflammatory cascade drive PCa aggressiveness. Hence, an integrative genomic approach may provide us with effective prognostic and diagnostic tools to manage inflammation-driven PCa progression. The aim of this study was to characterize inflammation-associated hub genes as predictors of PCa progression and investigate the effect of aberrant signaling of interleukin-1 receptor-associated kinases (IRAKs) on prostate tumorigenesis. WES and RNA-seq datasets from a total of 726 PCa patients, stratified into subtypes: primary PCa (PPC; n=493), metastatic PCa (MPC; n=150) and the castration-resistant PCa (CRPC; n=52) and neuroendocrine PCa (NEPC; n=30) from 3 cohort studies were acquired from cBioPortal. A bioinformatics pipeline was designed to analyze the datasets. Weighted gene co-expression network analysis (WGCNA) was used to construct gene co-expression networks and identify inflammation hub genes associated with PCa progression. The top inflammation hub genes, such as ILR1, TLRs, IRAK1-4, and NF-kB were validated. IRAK1-4 hub genes were found to be differentially expressed and genetically altered in NEPC and CRPC (&amp;gt;40%) compared to PPC and MPC (&amp;lt;20%) patients. GISTIC and MutSigCV were used to identify somatic mutations and copy-number alterations (CNAs). IRAK1 was found to have the highest CNA in CRPC and NEPC compared to MPC and PPC patients. Interestingly, no mutation was identified in CRPC and NEPC patients. The co-occurrence of IRAK1/4 CNVs and mRNA with AR, PI3K and AURKA signaling pathways suggest a role in PCa progression. To mechanistically investigate the role of IRAK1/4 signaling on PCa progression, CRISPR-mediated knockdown of IRAK1/4 was performed in PCa cell lines (LNCaP, PC3, and NCI-H660). mRNA and protein expression analyses were performed to determine changes in the levels of AR, PI3K/AKT, and AURKA between IRAK1/4-/- and IRAK1/4+/+ cell lines. Inhibition of IRAK1/4 significantly altered AR, PI3K, and AURKA signaling, leading to a reduction in PCa viability and metastatic potential. To investigate the effect of aberrant IRAK1/4 signaling on chemoresistance induction, IRAK+/+ PCa cells were treated with either TLR agonists or IRAK1/4 inhibitors and then treated with different anticancer drugs, such as Enzalutamide, Docetaxel, and Alisertib (AURKA inhibitor) dose-dependently. IRAK1/4-/- cells were more susceptible to anticancer drug treatment compared to IRAK1+/+ cells. The combination treatments were shown to be synergistic, thereby suppressing inflammation-induced chemoresistance. In conclusion, we have identified IRAK1/4 as a potential target for treating inflammation-driven prostate cancer progression. Citation Format: Saheed Oluwasina Oseni, Mirjana Pavlovic, James Hartmann, James Kumi-Diaka. Integrative genomic characterization and CRISPR-mediated gene editing studies identify IRAKs as novel therapeutic targets for inflammation-driven prostate tumorigenesis [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 995.

Systematic comparison of germline variant calling pipelines cross multiple next-generation sequencers

The development and innovation of next generation sequencing (NGS) and the subsequent analysis tools have gain popularity in scientific researches and clinical diagnostic applications. Hence, a systematic comparison of the sequencing platforms and variant calling pipelines could provide significant guidance to NGS-based scientific and clinical genomics. In this study, we compared the performance, concordance and operating efficiency of 27 combinations of sequencing platforms and variant calling pipelines, testing three variant calling pipelines—Genome Analysis Tool Kit HaplotypeCaller, Strelka2 and Samtools-Varscan2 for nine data sets for the NA12878 genome sequenced by different platforms including BGISEQ500, MGISEQ2000, HiSeq4000, NovaSeq and HiSeq Xten. For the variants calling performance of 12 combinations in WES datasets, all combinations displayed good performance in calling SNPs, with their F-scores entirely higher than 0.96, and their performance in calling INDELs varies from 0.75 to 0.91. And all 15 combinations in WGS datasets also manifested good performance, with F-scores in calling SNPs were entirely higher than 0.975 and their performance in calling INDELs varies from 0.71 to 0.93. All of these combinations manifested high concordance in variant identification, while the divergence of variants identification in WGS datasets were larger than that in WES datasets. We also down-sampled the original WES and WGS datasets at a series of gradient coverage across multiple platforms, then the variants calling period consumed by the three pipelines at each coverage were counted, respectively. For the GIAB datasets on both BGI and Illumina platforms, Strelka2 manifested its ultra-performance in detecting accuracy and processing efficiency compared with other two pipelines on each sequencing platform, which was recommended in the further promotion and application of next generation sequencing technology. The results of our researches will provide useful and comprehensive guidelines for personal or organizational researchers in reliable and consistent variants identification.

Detecting Expansions of Tandem Repeats in Cohorts Sequenced with Short-Read Sequencing Data

Repeat expansions cause more than 30 inherited disorders, predominantly neurogenetic. These can present with overlapping clinical phenotypes, making molecular diagnosis challenging. Single-gene or small-panel PCR-based methods can help to identify the precise genetic cause, but they can be slow and costly and often yield no result. Researchers are increasingly performing genomic analysis via whole-exome and whole-genome sequencing (WES and WGS) to diagnose genetic disorders. However, until recently, analysis protocols could not identify repeat expansions in these datasets. We developed exSTRa (expanded short tandem repeat algorithm), a method that uses either WES or WGS to identify repeat expansions. Performance of exSTRa was assessed in a simulation study. In addition, four retrospective cohorts of individuals with eleven different known repeat-expansion disorders were analyzed with exSTRa. We assessed results by comparing the findings to known disease status. Performance was also compared to three other analysis methods (ExpansionHunter, STRetch, and TREDPARSE), which were developed specifically for WGS data. Expansions in the assessed STR loci were successfully identified in WES and WGS datasets by all four methods with high specificity and sensitivity. Overall, exSTRa demonstrated more robust and superior performance for WES data than did the other three methods. We demonstrate that exSTRa can be effectively utilized as a screening tool for detecting repeat expansions in WES and WGS data, although the best performance would be produced by consensus calling, wherein at least two out of the four currently available screening methods call an expansion.

Secondary findings in 421 whole exome-sequenced Chinese children

BackgroundVariants with known or possible pathogenicity located in genes that are unrelated to primary disease conditions are defined as secondary findings. Secondary findings are not the primary targets of whole exome and genome sequencing (WES/WGS) assay but can be of great practical value in early disease prevention and intervention. The driving force for this study was to investigate the impact of racial difference and disease background on secondary findings. Here, we analyzed secondary findings frequencies in 421 whole exome-sequenced Chinese children who are phenotypically normal or bear congenital heart diseases/juvenile obesity. In total, 421 WES datasets were processed for potential deleterious variant screening. A reference gene list was defined according to the American College of Medical Genetics and Genomics (ACMG) recommendations for reporting secondary findings v2.0 (ACMG SF v2.0). The variant classification was performed according to the evidence-based guidelines recommended by the joint consensus of the ACMG and the Association for Molecular Pathology (AMP).ResultsAmong the 421 WES datasets, we identified 11 known/expected pathogenic variants in 12 individuals, accounting for 2.85% of our samples, which is much higher than the reported frequency in a Caucasian population. In conclusion, secondary findings are not so rare in Chinese children, which means that we should pay more attention to the clinical interpretation of sequencing results.