A somatic reference standard for cancer genome sequencing

Accelerated somatic mutation calling for whole-genome and whole-exome sequencing data from heterogenous tumor samples

Accurate detection of somatic mutations in DNA sequencing data is a fundamental prerequisite for cancer research. Previous analytical challenges were overcome by consensus mutation calling from four to five popular callers. This, however, increases the already nontrivial computing time from individual callers. Here, we launch MuSE 2, powered by multistep parallelization and efficient memory allocation, to resolve the computing time bottleneck. MuSE 2 speeds up 50 times more than MuSE 1 and eight to 80 times more than other popular callers. Our benchmark study suggests combining MuSE 2 and the recently accelerated Strelka2 achieves high efficiency and accuracy in analyzing large cancer genomic data sets.

The benefit of a complete reference genome for cancer structural variant analysis

The complexities of cancer genomes are becoming more easily interpreted due to advancements in sequencing technologies and improved bioinformatic analysis. Structural variants (SVs) represent an important subset of somatic events in tumors. While detection of SVs has been markedly improved by the development of long-read sequencing, somatic variant identification and annotation remains challenging. We hypothesized that use of a completed human reference genome (CHM13-T2T) would improve somatic SV calling. Our findings in a tumour/normal matched benchmark sample and two patient samples show that the CHM13-T2T improves SV detection and prioritization accuracy compared to GRCh38, with a notable reduction in false positive calls. We also overcame the lack of annotation resources for CHM13-T2T by lifting over CHM13-T2T-aligned reads to the GRCh38 genome, therefore combining both improved alignment and advanced annotations. In this process, we assessed the current SV benchmark set for COLO829/COLO829BL across four replicates sequenced at different centers with different long-read technologies. We discovered instability of this cell line across these replicates; 346 SVs (1.13%) were only discoverable in a single replicate. We identify 49 somatic SVs, which appear to be stable as they are consistently present across the four replicates. As such, we propose this consensus set as an updated benchmark for somatic SV calling and include both GRCh38 and CHM13-T2T coordinates in our benchmark. The benchmark is available at: 10.5281/zenodo.10819636 Our work demonstrates new approaches to optimize somatic SV prioritization in cancer with potential improvements in other genetic diseases.

SJPedPanel: A pan-cancer gene panel for childhood malignancies

Background

Large scale genomics projects have identified driver alterations for most childhood cancers that provide reliable biomarkers for clinical diagnosis and disease monitoring using targeted sequencing. However, there is lack of a comprehensive panel that matches the list of known driver genes. Here we fill this gap by developing SJPedPanel for childhood cancers.

Results

SJPedPanel covers 5,275 coding exons of 357 driver genes, 297 introns frequently involved in rearrangements that generate fusion oncoproteins, commonly amplified/deleted regions (e.g., MYCN for neuroblastoma, CDKN2A and PAX5 for B-/T-ALL, and SMARCB1 for AT/RT), and 7,590 polymorphism sites for interrogating tumors with aneuploidy, such as hyperdiploid and hypodiploid B-ALL or 17q gain neuroblastoma. We used driver alterations reported from an established real-time clinical genomics cohort (n=253) to validate this gene panel. Among the 485 pathogenic variants reported, our panel covered 417 variants (86%). For 90 rearrangements responsible for oncogenic fusions, our panel covered 74 events (82%). We re-sequenced 113 previously characterized clinical specimens at an average depth of 2,500X using SJPedPanel and recovered 354 (91%) of the 389 reported pathogenic variants. We then investigated the power of this panel in detecting mutations from specimens with low tumor purity (as low as 0.1%) using cell line-based dilution experiments and discovered that this gene panel enabled us to detect ∼80% variants with allele fraction of 0.2%, while the detection rate decreases to ∼50% when the allele fraction is 0.1%. We finally demonstrate its utility in disease monitoring on clinical specimens collected from AML patients in morphologic remission.

Conclusions

SJPedPanel enables the detection of clinically relevant genetic alterations including rearrangements responsible for subtype-defining fusions for childhood cancers by targeted sequencing of ∼0.15% of human genome. It will enhance the analysis of specimens with low tumor burdens for cancer monitoring and early detection.

Precise characterization of somatic complex structural variations from tumor/control paired long-read sequencing data with nanomonsv

We present our novel software, nanomonsv, for detecting somatic structural variations (SVs) using tumor and matched control long-read sequencing data with a single-base resolution. The current version of nanomonsv includes two detection modules, Canonical SV module, and Single breakend SV module. Using tumor/control paired long-read sequencing data from three cancer and their matched lymphoblastoid lines, we demonstrate that Canonical SV module can identify somatic SVs that can be captured by short-read technologies with higher precision and recall than existing methods. In addition, we have developed a workflow to classify mobile element insertions while elucidating their in-depth properties, such as 5' truncations, internal inversions, as well as source sites for 3' transductions. Furthermore, Single breakend SV module enables the detection of complex SVs that can only be identified by long-reads, such as SVs involving highly-repetitive centromeric sequences, and LINE1- and virus-mediated rearrangements. In summary, our approaches applied to cancer long-read sequencing data can reveal various features of somatic SVs and will lead to a better understanding of mutational processes and functional consequences of somatic SVs.

Simple combination of multiple somatic variant callers to increase accuracy

Publications comparing variant caller algorithms present discordant results with contradictory rankings. Caller performances are inconsistent and wide ranging, and dependent upon input data, application, parameter settings, and evaluation metric. With no single variant caller emerging as a superior standard, combinations or ensembles of variant callers have appeared in the literature. In this study, a whole genome somatic reference standard was used to derive principles to guide strategies for combining variant calls. Then, manually annotated variants called from the whole exome sequencing of a tumor were used to corroborate these general principles. Finally, we examined the ability of these principles to reduce noise in targeted sequencing.

Variant calling and benchmarking in an era of complete human genome sequences

Genetic variant calling from DNA sequencing has enabled understanding of germline variation in hundreds of thousands of humans. Sequencing technologies and variant-calling methods have advanced rapidly, routinely providing reliable variant calls in most of the human genome. We describe how advances in long reads, deep learning, de novo assembly and pangenomes have expanded access to variant calls in increasingly challenging, repetitive genomic regions, including medically relevant regions, and how new benchmark sets and benchmarking methods illuminate their strengths and limitations. Finally, we explore the possible future of more complete characterization of human genome variation in light of the recent completion of a telomere-to-telomere human genome reference assembly and human pangenomes, and we consider the innovations needed to benchmark their newly accessible repetitive regions and complex variants.

Towards routine chromosome-scale haplotype-resolved reconstruction in cancer genomics

Cancer genomes are highly complex and heterogeneous. The standard short-read sequencing and analytical methods are unable to provide the complete and precise base-level structural variant landscape of cancer genomes. In this work, we apply high-resolution long accurate HiFi and long-range Hi-C sequencing to the melanoma COLO829 cancer line. Also, we develop an efficient graph-based approach that processes these data types for chromosome-scale haplotype-resolved reconstruction to characterise the cancer precise structural variant landscape. Our method produces high-quality phased scaffolds on the chromosome level on three healthy samples and the COLO829 cancer line in less than half a day even in the absence of trio information, outperforming existing state-of-the-art methods. In the COLO829 cancer cell line, here we show that our method identifies and characterises precise somatic structural variant calls in important repeat elements that were missed in short-read-based call sets. Our method also finds the precise chromosome-level structural variant (germline and somatic) landscape with 19,956 insertions, 14,846 deletions, 421 duplications, 52 inversions and 498 translocations at the base resolution. Our simple pstools approach should facilitate better personalised diagnosis and disease management, including predicting therapeutic responses.

Heteroplasmic Mutant Load Differences in Mitochondrial DNA-Associated Leigh Syndrome

BACKGROUND

Mitochondrial DNA (mtDNA)-associated Leigh syndrome is influenced by mutant pathogenicity and corresponding heteroplasmic loads; however, the manner in which heteroplasmic mutant load affects patient phenotypes and the relationship between mutant types and heteroplasmic mutant loads remain unknown. We aimed to investigate the distribution of the mutant load of different mtDNA mutations in a single-center cohort.

METHODS

We used next-generation sequencing to confirm mtDNA mutations in 31 patients with Leigh syndrome. Subsequently, we counted the number of mtDNA reads to quantitatively analyze the heteroplasmic mutant load and categorize the patients according to the mtDNA mutations they harbored. Confirmed cases of mtDNA-associated Leigh syndrome were classified according to the mutations observed in six genes and 10 nucleotides.

RESULTS

Of the 31 patients with Leigh syndrome, 27 harbored known pathogenic mutations. We discovered that MT-ATP6 was the most commonly mutated gene (n = 13 patients), followed by MT-ND3 (n = 7) and MT-ND5 (n = 4). MT-ATP6 had a significantly higher mutant load than MT-ND3 and MT-ND5 (P < 0.001, each). By contrast, MT-ND5 had a significantly lower mutant load than MT-ND3 (P = 0.007). Notably, the mutation loads varied significantly among patients carrying the MT-ATP6, MT-ND3, and MT-ND5 mutations.

CONCLUSIONS

Our study illustrated the heteroplasmic diversity and phenotypic expression threshold of mutated mitochondrial genes in mtDNA-associated Leigh syndrome. The results provide promising insights into the genotype-phenotype correlation in mtDNA-associated Leigh syndrome that are expected to guide the development of tailored treatments for Leigh syndrome.

Accurate somatic variant detection using weakly supervised deep learning

Identification of somatic mutations in tumor samples is commonly based on statistical methods in combination with heuristic filters. Here we develop VarNet, an end-to-end deep learning approach for identification of somatic variants from aligned tumor and matched normal DNA reads. VarNet is trained using image representations of 4.6 million high-confidence somatic variants annotated in 356 tumor whole genomes. We benchmark VarNet across a range of publicly available datasets, demonstrating performance often exceeding current state-of-the-art methods. Overall, our results demonstrate how a scalable deep learning approach could augment and potentially supplant human engineered features and heuristic filters in somatic variant calling.

svaRetro and svaNUMT: modular packages for annotating retrotransposed transcripts and nuclear integration of mitochondrial DNA in genome sequencing data

Nuclear integration of mitochondrial genomes and retrocopied transcript insertion are biologically important but often-overlooked aspects of structural variant (SV) annotation. While tools for their detection exist, these typically rely on reanalysis of primary data using specialised detectors rather than leveraging calls from general purpose structural variant callers. Such reanalysis potentially leads to additional computational expense and does not take advantage of advances in general purpose structural variant calling. Here, we present svaRetro and svaNUMT; R packages that provide functions for annotating novel genomic events, such as nonreference retrocopied transcripts and nuclear integration of mitochondrial DNA. The packages were developed to work within the Bioconductor framework. We evaluate the performance of these packages to detect events using simulations and public benchmarking datasets, and annotate processed transcripts in a public structural variant database. svaRetro and svaNUMT provide modular, SV-caller agnostic tools for downstream annotation of structural variant calls.

A multi-platform reference for somatic structural variation detection

Accurate detection of somatic structural variation (SV) in cancer genomes remains a challenging problem. This is in part due to the lack of high-quality, gold-standard datasets that enable the benchmarking of experimental approaches and bioinformatic analysis pipelines. Here, we performed somatic SV analysis of the paired melanoma and normal lymphoblastoid COLO829 cell lines using four different sequencing technologies. Based on the evidence from multiple technologies combined with extensive experimental validation, we compiled a comprehensive set of carefully curated and validated somatic SVs, comprising all SV types. We demonstrate the utility of this resource by determining the SV detection performance as a function of tumor purity and sequence depth, highlighting the importance of assessing these parameters in cancer genomics projects. The truth somatic SV dataset as well as the underlying raw multi-platform sequencing data are freely available and are an important resource for community somatic benchmarking efforts.

Genotype-phenotype analysis of MT-ATP6-associated Leigh syndrome

OBJECTIVES

Mitochondrial DNA (mtDNA)-associated Leigh syndrome (LS) is characterized by maternal inheritance, and the heteroplasmic mutant load of mtDNA pathogenic variants is known to affect clinical phenotypes. Among mtDNA pathogenic variants, variants of the MT-ATP6 gene account for most of reported cases. In this report, we aimed to describe the clinical and genetic findings of MT-ATP6-associated LS patients diagnosed at a single tertiary institution in Korea.

METHODS

Thirteen patients with genetically confirmed MT-ATP6-associated LS were selected. We reviewed each patient's clinical findings, including general characteristics, biochemical parameters, brain MR images, muscle biopsy results, and heteroplasmic mutant load over a long-term follow-up period.

RESULTS

MT-ATP6-associated LS was of predominantly early onset (age <2 years), although we identified 2 late-onset (>60 months) LS patients. The heteroplasmic mutant load estimated by next-generation sequencing was 96%-100% in all nucleotide change groups. Compared with other forms of MT-ATP6-associated LS, the m.8993T>G point mutation elicited a significantly higher rate of symptom onset before 2 years of age. Brain MRI showed bilateral basal ganglia involvement in all patients, followed by cerebral atrophy, brainstem and thalamus involvement, and cerebellar atrophy. After follow-up (median 7.2 years, range 1.4 to 11.5 years), LS with m.8993T>G point mutations had a slightly more severe clinical progression compared with other forms of MT-ATP6-associated LS.

CONCLUSIONS

MT-ATP6-associated LS patients presented with a broad spectrum of clinical diagnoses and had a very high heteroplasmic mutant load. This study provides valuable data on MT-ATP6-associated LS that will inform subsequent studies on LS.

Consistency and reproducibility of large panel next-generation sequencing: Multi-laboratory assessment of somatic mutation detection on reference materials with mismatch repair and proofreading deficiency

INTRODUCTION

Clinical precision oncology increasingly relies on accurate genome-wide profiling using large panel next generation sequencing; however, difficulties in accurate and consistent detection of somatic mutation from individual platforms and pipelines remain an open question.

OBJECTIVES

To obtain paired tumor-normal reference materials that can be effectively constructed and interchangeable with clinical samples, and evaluate the performance of 56 panels under routine testing conditions based on the reference samples.

METHODS

Genes involved in mismatch repair and DNA proofreading were knocked down using the CRISPR-Cas9 technology to accumulate somatic mutations in a defined GM12878 cell line. They were used as reference materials to comprehensively evaluate the reproducibility and accuracy of detection results of oncopanels and explore the potential influencing factors.

RESULTS

In total, 14 paired tumor-normal reference DNA samples from engineered cell lines were prepared, and a reference dataset comprising 168 somatic mutations in a high-confidence region of 1.8 Mb were generated. For mutations with an allele frequency (AF) of more than 5% in reference samples, 56 panels collectively reported 1306 errors, including 729 false negatives (FNs), 179 false positives (FPs) and 398 reproducibility errors. The performance metric varied among panels with precision and recall ranging from 0.773 to 1 and 0.683 to 1, respectively. Incorrect and inadequate filtering accounted for a large proportion of false discovery (including FNs and FPs), while low-quality detection, cross-contamination and other sequencing errors during the wet bench process were other sources of FNs and FPs. In addition, low AF (<5%) considerably influenced the reproducibility and comparability among panels.

CONCLUSIONS

This study provided an integrated practice for developing reference standard to assess oncopanels in detecting somatic mutations and quantitatively revealed the source of detection errors. It will promote optimization, validation, and quality control among laboratories with potential applicability in clinical use.

Unscrambling cancer genomes via integrated analysis of structural variation and copy number

Complex somatic genomic rearrangements and copy number alterations are hallmarks of nearly all cancers. We have developed an algorithm, LINX, to aid interpretation of structural variant and copy number data derived from short-read, whole-genome sequencing. LINX classifies raw structural variant calls into distinct events and predicts their effect on the local structure of the derivative chromosome and the functional impact on affected genes. Visualizations facilitate further investigation of complex rearrangements. LINX allows insights into a diverse range of structural variation events and can reliably detect pathogenic rearrangements, including gene fusions, immunoglobulin enhancer rearrangements, intragenic deletions, and duplications. Uniquely, LINX also predicts chained fusions that we demonstrate account for 13% of clinically relevant oncogenic fusions. LINX also reports a class of inactivation events that we term homozygous disruptions that may be a driver mutation in up to 9% of tumors and may frequently affect PTEN, TP53, and RB1.

Achieving robust somatic mutation detection with deep learning models derived from reference data sets of a cancer sample

BACKGROUND

Accurate detection of somatic mutations is challenging but critical in understanding cancer formation, progression, and treatment. We recently proposed NeuSomatic, the first deep convolutional neural network-based somatic mutation detection approach, and demonstrated performance advantages on in silico data.

RESULTS

In this study, we use the first comprehensive and well-characterized somatic reference data sets from the SEQC2 consortium to investigate best practices for using a deep learning framework in cancer mutation detection. Using the high-confidence somatic mutations established for a cancer cell line by the consortium, we identify the best strategy for building robust models on multiple data sets derived from samples representing real scenarios, for example, a model trained on a combination of real and spike-in mutations had the highest average performance.

CONCLUSIONS

The strategy identified in our study achieved high robustness across multiple sequencing technologies for fresh and FFPE DNA input, varying tumor/normal purities, and different coverages, with significant superiority over conventional detection approaches in general, as well as in challenging situations such as low coverage, low variant allele frequency, DNA damage, and difficult genomic regions.

Association Between Epilepsy and Leigh Syndrome With MT-ND3 Mutation, Particularly the m.10191T>C Point Mutation

Background and Purpose: Recent advances in molecular genetic testing have led to a rapid increase in the understanding of the genetics of Leigh syndrome. Several studies have suggested that Leigh syndrome with MT-ND3 mutation is strongly associated with epilepsy. This study focused on the epilepsy-related characteristics of Leigh syndrome with MT-ND3 mutation identified in a single tertiary hospital in South Korea. Methods: We selected 31 patients with mitochondrial DNA (mtDNA) mutations who were genetically diagnosed with mtDNA-associated Leigh syndrome. Among them, seven patients with MT-ND3 mutations were detected. We reviewed various clinical findings such as laboratory findings, brain images, electroencephalography data, seizure types, seizure frequency, antiepileptic drug use history, and current seizure status. Results: The nucleotide changes in the seven patients with the Leigh syndrome with MT-ND3 mutation were divided into two groups: m.10191T>C and m.10158T>C. Six of the seven patients were found to have the m.10191T>C mutations. The median value of the mutant load was 82.5%, ranging from 57.9 to 93.6%. No particular tendency was observed for the first symptom or seizure onset or mutant load. The six patients with the m.10191T>C mutation were diagnosed with epilepsy. Three of these patients were diagnosed with Lennox-Gastaut syndrome (LGS). Conclusion: We reported a very strong association between epilepsy and MT-ND3 mutation in Leigh syndrome, particularly the m.10191T>C mutation. The possibility of an association between the epilepsy phenotype of the m.10191T>C mutation and LGS was noted.

FDA-led consortium studies advance quality control of targeted next generation sequencing assays for precision oncology

Cancer is the second leading cause of mortality worldwide despite tremendous advances in treatment. The promise of precision oncology depends on accurate characterization of tumor mutations and subsequent therapy selection. The lack of tumor reference samples along with the associated next generation sequencing (NGS) technical assessments has hindered the development of NGS assays and the realization of benefits for precision oncology. The summarized results and recommendations of several seminal SEQC2 studies along with a vision of the changing landscape of precision oncology and anticipated next steps by the SEQC2 consortium are reported. Importantly, these studies utilized a new robust reference sample material which was developed and constructed to support multiple DNA and RNA-based NGS assay studies. These studies focused on a wide variety of precision oncology assay scenarios and provided guidelines for standardized analyses and best practice recommendations. The evolving landscape of precision oncology requires insights into critical factors supporting the sensitivity and reproducibility of clinical NGS assays for continued improvement in patient outcomes. Persistent development of robust reference materials, quantitative performance metrics, and actionable data analysis recommendations are needed. This series of SEQC2 studies serve to advance NGS-based assays for precision oncology and support regulatory science endeavors.

Advancing NGS quality control to enable measurement of actionable mutations in circulating tumor DNA

The primary objective of the FDA-led Sequencing and Quality Control Phase 2 (SEQC2) project is to develop standard analysis protocols and quality control metrics for use in DNA testing to enhance scientific research and precision medicine. This study reports a targeted next-generation sequencing (NGS) method that will enable more accurate detection of actionable mutations in circulating tumor DNA (ctDNA) clinical specimens. To accomplish this, a synthetic internal standard spike-in was designed for each actionable mutation target, suitable for use in NGS following hybrid capture enrichment and unique molecular index (UMI) or non-UMI library preparation. When mixed with contrived ctDNA reference samples, internal standards enabled calculation of technical error rate, limit of blank, and limit of detection for each variant at each nucleotide position in each sample. True-positive mutations with variant allele fraction too low for detection by current practice were detected with this method, thereby increasing sensitivity.

Cross-Site Concordance Evaluation of Tumor DNA and RNA Sequencing Platforms for the CIMAC-CIDC Network

PURPOSE

Whole-exome (WES) and RNA sequencing (RNA-seq) are key components of cancer immunogenomic analyses. To evaluate the consistency of tumor WES and RNA-seq profiling platforms across different centers, the Cancer Immune Monitoring and Analysis Centers (CIMAC) and the Cancer Immunologic Data Commons (CIDC) conducted a systematic harmonization study.

EXPERIMENTAL DESIGN

DNA and RNA were centrally extracted from fresh frozen and formalin-fixed paraffin-embedded non-small cell lung carcinoma tumors and distributed to three centers for WES and RNA-seq profiling. In addition, two 10-plex HapMap cell line pools with known mutations were used to evaluate the accuracy of the WES platforms.

RESULTS

The WES platforms achieved high precision (> 0.98) and recall (> 0.87) on the HapMap pools when evaluated on loci using > 50× common coverage. Nonsynonymous mutations clustered by tumor sample, achieving an index of specific agreement above 0.67 among replicates, centers, and sample processing. A DV200 > 24% for RNA, as a putative presequencing RNA quality control (QC) metric, was found to be a reliable threshold for generating consistent expression readouts in RNA-seq and NanoString data. MedTIN > 30 was likewise assessed as a reliable RNA-seq QC metric, above which samples from the same tumor across replicates, centers, and sample processing runs could be robustly clustered and HLA typing, immune infiltration, and immune repertoire inference could be performed.

CONCLUSIONS

The CIMAC collaborating laboratory platforms effectively generated consistent WES and RNA-seq data and enable robust cross-trial comparisons and meta-analyses of highly complex immuno-oncology biomarker data across the NCI CIMAC-CIDC Network.

Establishing community reference samples, data and call sets for benchmarking cancer mutation detection using whole-genome sequencing

The lack of samples for generating standardized DNA datasets for setting up a sequencing pipeline or benchmarking the performance of different algorithms limits the implementation and uptake of cancer genomics. Here, we describe reference call sets obtained from paired tumor-normal genomic DNA (gDNA) samples derived from a breast cancer cell line-which is highly heterogeneous, with an aneuploid genome, and enriched in somatic alterations-and a matched lymphoblastoid cell line. We partially validated both somatic mutations and germline variants in these call sets via whole-exome sequencing (WES) with different sequencing platforms and targeted sequencing with >2,000-fold coverage, spanning 82% of genomic regions with high confidence. Although the gDNA reference samples are not representative of primary cancer cells from a clinical sample, when setting up a sequencing pipeline, they not only minimize potential biases from technologies, assays and informatics but also provide a unique resource for benchmarking 'tumor-only' or 'matched tumor-normal' analyses.

CRISPAltRations: a validated cloud-based approach for interrogation of double-strand break repair mediated by CRISPR genome editing

CRISPR systems enable targeted genome editing in a wide variety of organisms by introducing single- or double-strand DNA breaks, which are repaired using endogenous molecular pathways. Characterization of on- and off-target editing events from CRISPR proteins can be evaluated using targeted genome resequencing. We characterized DNA repair fingerprints that result from non-homologous end joining (NHEJ) after double-stranded breaks (DSBs) were introduced by Cas9 or Cas12a for >500 paired treatment/control experiments. We found that building biological understanding of the repair into a novel analysis tool (CRISPAltRations) improved the quality of the results. We validated our software using simulated, targeted amplicon sequencing data (11 guide RNAs [gRNAs] and 603 on- and off-target locations) and demonstrated that CRISPAltRations outperforms other publicly available software tools in accurately annotating CRISPR-associated indels and homology-directed repair (HDR) events. We enable non-bioinformaticians to use CRISPAltRations by developing a web-accessible, cloud-hosted deployment, which allows rapid batch processing of samples in a graphical user interface (GUI) and complies with HIPAA security standards. By ensuring that our software is thoroughly tested, version controlled, and supported with a user interface (UI), we enable resequencing analysis of CRISPR genome editing experiments to researchers no matter their skill in bioinformatics.

The Application of Control Materials for Ongoing Quality Management of Next-Generation Sequencing in a Clinical Genetic Laboratory

Next-generation sequencing (NGS) has played an important role in detecting genetic variants with pathologic and therapeutic potential. The advantages of NGS, such as high-throughput sequencing capacity and massively parallel sequencing, have a significant impact on realization of genetic profiling in clinical genetic laboratories. These changes have enabled clinicians to execute precision medicine in diagnosis, prognosis, and treatment for patients. However, to adapt targeted gene panels in diagnostic use, analytical validation and ongoing quality control should be implemented and applied with both practical guidelines and appropriate control materials. Several guidelines for NGS quality control recommend usage of control materials such as HapMap cell lines, synthetic DNA fragments, and genetically characterized cell lines; however, specifications or applications of such usage are insufficient to guideline method development. This review focuses on what factors should be considered before control material selection for NGS assay and practical methods of how they could be developed in clinical genetic laboratories. This review also provides the detailed sources of critical information related to control materials.

A verified genomic reference sample for assessing performance of cancer panels detecting small variants of low allele frequency

BACKGROUND

Oncopanel genomic testing, which identifies important somatic variants, is increasingly common in medical practice and especially in clinical trials. Currently, there is a paucity of reliable genomic reference samples having a suitably large number of pre-identified variants for properly assessing oncopanel assay analytical quality and performance. The FDA-led Sequencing and Quality Control Phase 2 (SEQC2) consortium analyze ten diverse cancer cell lines individually and their pool, termed Sample A, to develop a reference sample with suitably large numbers of coding positions with known (variant) positives and negatives for properly evaluating oncopanel analytical performance.

RESULTS

In reference Sample A, we identify more than 40,000 variants down to 1% allele frequency with more than 25,000 variants having less than 20% allele frequency with 1653 variants in COSMIC-related genes. This is 5-100× more than existing commercially available samples. We also identify an unprecedented number of negative positions in coding regions, allowing statistical rigor in assessing limit-of-detection, sensitivity, and precision. Over 300 loci are randomly selected and independently verified via droplet digital PCR with 100% concordance. Agilent normal reference Sample B can be admixed with Sample A to create new samples with a similar number of known variants at much lower allele frequency than what exists in Sample A natively, including known variants having allele frequency of 0.02%, a range suitable for assessing liquid biopsy panels.

CONCLUSION

These new reference samples and their admixtures provide superior capability for performing oncopanel quality control, analytical accuracy, and validation for small to large oncopanels and liquid biopsy assays.

SequencErr: measuring and suppressing sequencer errors in next-generation sequencing data

Background

There is currently no method to precisely measure the errors that occur in the sequencing instrument/sequencer, which is critical for next-generation sequencing applications aimed at discovering the genetic makeup of heterogeneous cellular populations.

Results

We propose a novel computational method, SequencErr, to address this challenge by measuring the base correspondence between overlapping regions in forward and reverse reads. An analysis of 3777 public datasets from 75 research institutions in 18 countries revealed the sequencer error rate to be ~ 10 per million (pm) and 1.4% of sequencers and 2.7% of flow cells have error rates > 100 pm. At the flow cell level, error rates are elevated in the bottom surfaces and > 90% of HiSeq and NovaSeq flow cells have at least one outlier error-prone tile. By sequencing a common DNA library on different sequencers, we demonstrate that sequencers with high error rates have reduced overall sequencing accuracy, and removal of outlier error-prone tiles improves sequencing accuracy. We demonstrate that SequencErr can reveal novel insights relative to the popular quality control method FastQC and achieve a 10-fold lower error rate than popular error correction methods including Lighter and Musket.

Conclusions

Our study reveals novel insights into the nature of DNA sequencing errors incurred on DNA sequencers. Our method can be used to assess, calibrate, and monitor sequencer accuracy, and to computationally suppress sequencer errors in existing datasets.

Whole-slide laser microdissection for tumour enrichment

The practical application of genome-scale technologies to precision oncology research requires flexible tissue processing strategies that can be used to differentially select both tumour and normal cell populations from formalin-fixed, paraffin-embedded tissues. As tumour sequencing scales towards clinical implementation, practical difficulties in scheduling and obtaining fresh tissue biopsies at scale, including blood samples as surrogates for matched 'normal' DNA, have focused attention on the use of formalin-preserved clinical samples collected routinely for diagnostic purposes. In practice, such samples often contain both tumour and normal cells which, if correctly partitioned, could be used to profile both tumour and normal genomes, thus identifying somatic alterations. Here we report a semi-automated method for laser microdissecting entire slide-mounted tissue sections to enrich for cells of interest with sufficient yield for whole genome and transcriptome sequencing. Using this method, we demonstrated enrichment of tumour material from mixed tumour-normal samples by up to 67%. Leveraging new methods that allow for the extraction of high-quality nucleic acids from small amounts of formalin-fixed tissues, we further showed that the method was successful in yielding sequence data of sufficient quality for use in BC Cancer's Personalized OncoGenomics (POG) program. © 2020 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

A Distributed Whole Genome Sequencing Benchmark Study

Population sequencing often requires collaboration across a distributed network of sequencing centers for the timely processing of thousands of samples. In such massive efforts, it is important that participating scientists can be confident that the accuracy of the sequence data produced is not affected by which center generates the data. A study was conducted across three established sequencing centers, located in Montreal, Toronto, and Vancouver, constituting Canada's Genomics Enterprise (www.cgen.ca). Whole genome sequencing was performed at each center, on three genomic DNA replicates from three well-characterized cell lines. Secondary analysis pipelines employed by each site were applied to sequence data from each of the sites, resulting in three datasets for each of four variables (cell line, replicate, sequencing center, and analysis pipeline), for a total of 81 datasets. These datasets were each assessed according to multiple quality metrics including concordance with benchmark variant truth sets to assess consistent quality across all three conditions for each variable. Three-way concordance analysis of variants across conditions for each variable was performed. Our results showed that the variant concordance between datasets differing only by sequencing center was similar to the concordance for datasets differing only by replicate, using the same analysis pipeline. We also showed that the statistically significant differences between datasets result from the analysis pipeline used, which can be unified and updated as new approaches become available. We conclude that genome sequencing projects can rely on the quality and reproducibility of aggregate data generated across a network of distributed sites.

DeepSSV: detecting somatic small variants in paired tumor and normal sequencing data with convolutional neural network

It is of considerable interest to detect somatic mutations in paired tumor and normal sequencing data. A number of callers that are based on statistical or machine learning approaches have been developed to detect somatic small variants. However, they take into consideration only limited information about the reference and potential variant allele in both tumor and normal samples at a candidate somatic site. Also, they differ in how biological and technological noises are addressed. Hence, they are expected to produce divergent outputs. To overcome the drawbacks of existing somatic callers, we develop a deep learning-based tool called DeepSSV, which employs a convolutional neural network (CNN) model to learn increasingly abstract feature representations from the raw data in higher feature layers. DeepSSV creates a spatially oriented representation of read alignments around the candidate somatic sites adapted for the convolutional architecture, which enables it to expand to effectively gather scattered evidence. Moreover, DeepSSV incorporates the mapping information of both reference allele-supporting and variant allele-supporting reads in the tumor and normal samples at a genomic site that are readily available in the pileup format file. Together, the CNN model can process the whole alignment information. Such representational richness allows the model to capture the dependencies in the sequence and identify context-based sequencing artifacts. We fitted the model on ground truth somatic mutations and did benchmarking experiments on simulated and real tumors. The benchmarking results demonstrate that DeepSSV outperforms its state-of-the-art competitors in overall F1 score.

Whole genome, transcriptome and methylome profiling enhances actionable target discovery in high-risk pediatric cancer

The Zero Childhood Cancer Program is a precision medicine program to benefit children with poor-outcome, rare, relapsed or refractory cancer. Using tumor and germline whole genome sequencing (WGS) and RNA sequencing (RNAseq) across 252 tumors from high-risk pediatric patients with cancer, we identified 968 reportable molecular aberrations (39.9% in WGS and RNAseq, 35.1% in WGS only and 25.0% in RNAseq only). Of these patients, 93.7% had at least one germline or somatic aberration, 71.4% had therapeutic targets and 5.2% had a change in diagnosis. WGS identified pathogenic cancer-predisposing variants in 16.2% of patients. In 76 central nervous system tumors, methylome analysis confirmed diagnosis in 71.1% of patients and contributed to a change of diagnosis in two patients (2.6%). To date, 43 patients have received a recommended therapy, 38 of whom could be evaluated, with 31% showing objective evidence of clinical benefit. Comprehensive molecular profiling resolved the molecular basis of virtually all high-risk cancers, leading to clinical benefit in some patients.

SICaRiO: short indel call filtering with boosting

Despite impressive improvement in the next-generation sequencing technology, reliable detection of indels is still a difficult endeavour. Recognition of true indels is of prime importance in many applications, such as personalized health care, disease genomics and population genetics. Recently, advanced machine learning techniques have been successfully applied to classification problems with large-scale data. In this paper, we present SICaRiO, a gradient boosting classifier for the reliable detection of true indels, trained with the gold-standard dataset from 'Genome in a Bottle' (GIAB) consortium. Our filtering scheme significantly improves the performance of each variant calling pipeline used in GIAB and beyond. SICaRiO uses genomic features that can be computed from publicly available resources, i.e. it does not require sequencing pipeline-specific information (e.g. read depth). This study also sheds lights on prior genomic contexts responsible for the erroneous calling of indels made by sequencing pipelines. We have compared prediction difficulty for three categories of indels over different sequencing pipelines. We have also ranked genomic features according to their predictivity in determining false positives.

The regulatory landscape of precision oncology laboratory medicine in the United States - Perspective on the past 5 years and considerations for future regulation

The regulatory landscape for precision oncology in the United States is complicated, with multiple governmental regulatory agencies with different scopes of jurisdiction. Several regulatory proposals have been introduced since the Food and Drug Administration released a draft guidance to regulate laboratory-developed tests in 2014. Key aspects of the most recent proposals and discussion of central arguments related to the regulation of precision oncology laboratory tests provides insight to stakeholders for future discussions related to regulation of laboratory tests.

SomaticCombiner: improving the performance of somatic variant calling based on evaluation tests and a consensus approach

It is challenging to identify somatic variants from high-throughput sequence reads due to tumor heterogeneity, sub-clonality, and sequencing artifacts. In this study, we evaluated the performance of eight primary somatic variant callers and multiple ensemble methods using both real and synthetic whole-genome sequencing, whole-exome sequencing, and deep targeted sequencing datasets with the NA12878 cell line. The test results showed that a simple consensus approach can significantly improve performance even with a limited number of callers and is more robust and stable than machine learning based ensemble approaches. To fully exploit the multi-callers, we also developed a software package, SomaticCombiner, that can combine multiple callers and integrates a new variant allelic frequency (VAF) adaptive majority voting approach, which can maintain sensitive detection for variants with low VAFs.

Single-cell sequencing of genomic DNA resolves sub-clonal heterogeneity in a melanoma cell line

We performed shallow single-cell sequencing of genomic DNA across 1475 cells from a cell-line, COLO829, to resolve overall complexity and clonality. This melanoma tumor-line has been previously characterized by multiple technologies and is a benchmark for evaluating somatic alterations. In some of these studies, COLO829 has shown conflicting and/or indeterminate copy number and, thus, single-cell sequencing provides a tool for gaining insight. Following shallow single-cell sequencing, we first identified at least four major sub-clones by discriminant analysis of principal components of single-cell copy number data. Based on clustering, break-point and loss of heterozygosity analysis of aggregated data from sub-clones, we identified distinct hallmark events that were validated within bulk sequencing and spectral karyotyping. In summary, COLO829 exhibits a classical Dutrillaux’s monosomic/trisomic pattern of karyotype evolution with endoreduplication, where consistent sub-clones emerge from the loss/gain of abnormal chromosomes. Overall, our results demonstrate how shallow copy number profiling can uncover hidden biological insights.

Through shallow single-cell sequencing of genomic DNA followed by clustering analysis, Velazquez-Villarreal et al. reveal sub-clones of the melanoma cell line COLO829 and further identify and validate chromosome translocations and copy number changes. This study illustrates how copy number variation analysis can provide insights into cancer cell heterogeneity.

PRESM: personalized reference editor for somatic mutation discovery in cancer genomics

MOTIVATION

Accurate detection of somatic mutations is a crucial step toward understanding cancer. Various tools have been developed to detect somatic mutations from cancer genome sequencing data by mapping reads to a universal reference genome and inferring likelihoods from complex statistical models. However, read mapping is frequently obstructed by mismatches between germline and somatic mutations on a read and the reference genome. Previous attempts to develop personalized genome tools are not compatible with downstream statistical models for somatic mutation detection.

RESULTS

We present PRESM, a tool that builds personalized reference genomes by integrating germline mutations into the reference genome. The aforementioned obstacle is circumvented by using a two-step germline substitution procedure, maintaining positional fidelity using an innovative workaround. Reads derived from tumor tissue can be positioned more accurately along a personalized reference than a universal reference due to the reduced genetic distance between the subject (tumor genome) and the target (the personalized genome). Application of PRESM's personalized genome reduced false-positive (FP) somatic mutation calls by as much as 55.5%, and facilitated the discovery of a novel somatic point mutation on a germline insertion in PDE1A, a phosphodiesterase associated with melanoma. Moreover, all improvements in calling accuracy were achieved without parameter optimization, as PRESM itself is parameter-free. Hence, similar increases in read mapping and decreases in the FP rate will persist when PRESM-built genomes are applied to any user-provided dataset.

AVAILABILITY AND IMPLEMENTATION

The software is available at https://github.com/precisionomics/PRESM.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Deep whole-genome sequencing of 3 cancer cell lines on 2 sequencing platforms

To test the performance of a new sequencing platform, develop an updated somatic calling pipeline and establish a reference for future benchmarking experiments, we performed whole-genome sequencing of 3 common cancer cell lines (COLO-829, HCC-1143 and HCC-1187) along with their matched normal cell lines to great sequencing depths (up to 278x coverage) on both Illumina HiSeqX and NovaSeq sequencing instruments. Somatic calling was generally consistent between the two platforms despite minor differences at the read level. We designed and implemented a novel pipeline for the analysis of tumor-normal samples, using multiple variant callers. We show that coupled with a high-confidence filtering strategy, the use of combination of tools improves the accuracy of somatic variant calling. We also demonstrate the utility of the dataset by creating an artificial purity ladder to evaluate the somatic pipeline and benchmark methods for estimating purity and ploidy from tumor-normal pairs. The data and results of the pipeline are made accessible to the cancer genomics community.

Joint detection of germline and somatic copy number events in matched tumor-normal sample pairs

MOTIVATION

Whole-genome sequencing (WGS) of tumor-normal sample pairs is a powerful approach for comprehensively characterizing germline copy number variations (CNVs) and somatic copy number alterations (SCNAs) in cancer research and clinical practice. Existing computational approaches for detecting copy number events cannot detect germline CNVs and SCNAs simultaneously, and yield low accuracy for SCNAs.

RESULTS

In this study, we developed TumorCNV, a novel approach for jointly detecting germline CNVs and SCNAs from WGS data of the matched tumor-normal sample pair. We compared TumorCNV with existing copy number event detection approaches using the simulated data and real data for the COLO-829 melanoma cell line. The experimental results showed that TumorCNV achieved superior performance than existing approaches.

AVAILABILITY AND IMPLEMENTATION

The software TumorCNV is implemented using a combination of Java and R, and it is freely available from the website at https://github.com/yongzhuang/TumorCNV.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Rare variant phasing using paired tumor:normal sequence data

Background

In standard high throughput sequencing analysis, genetic variants are not assigned to a homologous chromosome of origin. This process, called haplotype phasing, can reveal information important for understanding the relationship between genetic variants and biological phenotypes. For example, in genes that carry multiple heterozygous missense variants, phasing resolves whether one or both gene copies are altered. Here, we present a novel approach to phasing variants that takes advantage of unique properties of paired tumor:normal sequencing data from cancer studies.

Results

VAF phasing uses changes in variant allele frequency (VAF) between tumor and normal samples in regions of somatic chromosomal gain or loss to phase germline variants. We apply VAF phasing to 6180 samples from the Cancer Genome Atlas (TCGA) and demonstrate that our method is highly concordant with other standard phasing methods, and can phase an average of 33% more variants than other read-backed phasing methods. Using variant annotation tools designed to score gene haplotypes, we find a suggestive association between carrying multiple missense variants in a single copy of a cancer predisposition gene and earlier age of cancer diagnosis.

Conclusions

VAF phasing exploits unique properties of tumor genomes to increase the number of germline variants that can be phased over standard read-backed methods in paired tumor:normal samples. Our phase-informed association testing results call attention to the need to develop more tools for assessing the joint effect of multiple genetic variants.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-2753-1) contains supplementary material, which is available to authorized users.

Rearrangement bursts generate canonical gene fusions in bone and soft tissue tumors

Sarcomas are cancers of the bone and soft tissue often defined by gene fusions. Ewing sarcoma involves fusions between EWSR1, a gene encoding an RNA binding protein, and E26 transformation-specific (ETS) transcription factors. We explored how and when EWSR1-ETS fusions arise by studying the whole genomes of Ewing sarcomas. In 52 of 124 (42%) of tumors, the fusion gene arises by a sudden burst of complex, loop-like rearrangements, a process called chromoplexy, rather than by simple reciprocal translocations. These loops always contained the disease-defining fusion at the center, but they disrupted multiple additional genes. The loops occurred preferentially in early replicating and transcriptionally active genomic regions. Similar loops forming canonical fusions were found in three other sarcoma types. Chromoplexy-generated fusions appear to be associated with an aggressive form of Ewing sarcoma. These loops arise early, giving rise to both primary and relapse Ewing sarcoma tumors, which can continue to evolve in parallel.

Complexity of Delivering Precision Medicine: Opportunities and Challenges

Precision medicine has emerged as a tool to match patients with the appropriate treatment based on the precise molecular features of an individual patient's tumor. Although examples of targeted therapies exist resulting in dramatic improvements in patient outcomes, comprehensive genomic profiling of tumors has also demonstrated the incredible complexity of molecular alterations in tissue and blood. These sequencing methods provide opportunities to study the landscape of tumors at baseline and serially in response to treatment. These tools also serve as important biomarkers to detect resistance to treatment and determine higher likelihood of responding to particular treatments, such as immune checkpoint blockade. Federally funded and publicly available data repositories have emerged as mechanisms for data sharing. In addition, novel clinical trials are emerging to develop new ways of incorporating molecular matched therapy into clinical trials. Various challenges to delivery of precision oncology include understanding the complexity of advanced tumors based on evolving "omics" and treatment resistance. For physicians, determining when and how to incorporate genetic and molecular tools into clinic in a cost-effective manner is critical. Finally, we discuss the importance of well-designed prospective clinical trials, biomarkers such as liquid biopsies, the use of multidisciplinary tumor boards, and data sharing as evidence-based medicine tools to optimally study and deliver precision oncology to our patients.

A review of somatic single nucleotide variant calling algorithms for next-generation sequencing data

Detection of somatic mutations holds great potential in cancer treatment and has been a very active research field in the past few years, especially since the breakthrough of the next-generation sequencing technology. A collection of variant calling pipelines have been developed with different underlying models, filters, input data requirements, and targeted applications. This review aims to enumerate these unique features of the state-of-the-art variant callers, in the hope to provide a practical guide for selecting the appropriate pipeline for specific applications. We will focus on the detection of somatic single nucleotide variants, ranging from traditional variant callers based on whole genome or exome sequencing of paired tumor-normal samples to recent low-frequency variant callers designed for targeted sequencing protocols with unique molecular identifiers. The variant callers have been extensively benchmarked with inconsistent performances across these studies. We will review the reference materials, datasets, and performance metrics that have been used in the benchmarking studies. In the end, we will discuss emerging trends and future directions of the variant calling algorithms.

Whole Genome Library Construction for Next Generation Sequencing

With the rapid evolution of genomics technologies over the past decade, whole genome sequencing (WGS) has become an increasingly accessible tool in biomedical research. WGS applications include analysis of genomic DNA from single individuals, multiple related family members, and tumor/normal samples from the same patient in the context of oncology. A number of different modalities are available for performing WGS; this chapter focuses on wet lab library construction procedures for complex short insert WGS libraries using the KAPA Hyper Prep Kit (Kapa Biosystems), and includes a discussion of appropriate quality control measures for sequencing on the Illumina HiSeq2000 platform. Additional modifications to the protocol for long insert WGS library construction, to assess structural alterations and copy number changes, are also described.

Functional Mutations Form at CTCF-Cohesin Binding Sites in Melanoma Due to Uneven Nucleotide Excision Repair across the Motif

CTCF binding sites are frequently mutated in cancer, but how these mutations accumulate and whether they broadly perturb CTCF binding are not well understood. Here, we report that skin cancers exhibit a highly specific asymmetric mutation pattern within CTCF motifs attributable to ultraviolet irradiation and differential nucleotide excision repair (NER). CTCF binding site mutations form independently of replication timing and are enriched at sites of CTCF/cohesin complex binding, suggesting a role for cohesin in stabilizing CTCF-DNA binding and impairing NER. Performing CTCF ChIP-seq in a melanoma cell line, we show CTCF binding site mutations to be functional by demonstrating allele-specific reduction of CTCF binding to mutant alleles. While topologically associating domains with mutated CTCF anchors in melanoma contain differentially expressed cancer-associated genes, CTCF motif mutations appear generally under neutral selection. However, the frequency and potential functional impact of such mutations in melanoma highlights the need to consider their impact on cellular phenotype in individual genomes.