Somatic Structural Variants Research Articles

Whole-Genome Sequencing Reveals Diverse Models of Structural Variations in Esophageal Squamous Cell Carcinoma

Comprehensive identification of somatic structural variations (SVs) and understanding their mutational mechanisms in cancer might contribute to understanding biological differences and help to identify new therapeutic targets. Unfortunately, characterization of complex SVs across the whole genome and the mutational mechanisms underlying esophageal squamous cell carcinoma (ESCC) is largely unclear. To define a comprehensive catalog of somatic SVs, affected target genes, and their underlying mechanisms in ESCC, we re-analyzed whole-genome sequencing (WGS) data from 31 ESCCs using Meerkat algorithm to predict somatic SVs and Patchwork to determine copy-number changes. We found deletions and translocations with NHEJ and alt-EJ signature as the dominant SV types, and 16% of deletions were complex deletions. SVs frequently led to disruption of cancer-associated genes (e.g., CDKN2A and NOTCH1) with different mutational mechanisms. Moreover, chromothripsis, kataegis, and breakage-fusion-bridge (BFB) were identified as contributing to locally mis-arranged chromosomes that occurred in 55% of ESCCs. These genomic catastrophes led to amplification of oncogene through chromothripsis-derived double-minute chromosome formation (e.g., FGFR1 and LETM2) or BFB-affected chromosomes (e.g., CCND1, EGFR, ERBB2, MMPs, and MYC), with approximately 30% of ESCCs harboring BFB-derived CCND1 amplification. Furthermore, analyses of copy-number alterations reveal high frequency of whole-genome duplication (WGD) and recurrent focal amplification of CDCA7 that might act as a potential oncogene in ESCC. Our findings reveal molecular defects such as chromothripsis and BFB in malignant transformation of ESCCs and demonstrate diverse models of SVs-derived target genes in ESCCs. These genome-wide SV profiles and their underlying mechanisms provide preventive, diagnostic, and therapeutic implications for ESCCs.

Clincial and Genetic Landscapes Differ Between IGH-CRLF2 and P2RY8-CRLF2 Acute Lymphoblastic Leukaemia

Deregulated expression of the type I cytokine receptor, cytokine receptor-like factor 2 (CRLF2 -d), is observed in 5% of B-cell precursor acute lymphoblastic leukaemia (BCP-ALL). It occurs via three known aberrations; a cryptic reciprocal translocation with the immunoglobulin heavy chain locus (IGH); an interstitial deletion within PAR1, resulting in the P2RY8-CRLF2 fusion; rare but recurrent CRLF2 mutation (F323C). All three result in overexpression of CRLF2, however alone it is insufficient to cause overt leukaemia. Mutations of the Janus kinase (JAK) family genes and the IL7 receptor, are recurrent (50% of CRLF2-d) and together with CRLF2 -d result in transformation of mouse BaF3 cells. We noted that outcome data from different study groups were inconsistent, with patients being classified as either high or intermediate risk. Thus, in this largest study to date, our aim was to acertain whether differences in clinical and/or genetic features between patients with CRLF2-d and involvement of either IGH or P2RY8 may define them as different disease entities and partly explain the outcome heterogeneity.Among 160 CRLF2-d patients, chromosomal analysis confirmed a high incidence of additional somatic copies of chromosomes X (39/88, 44%) and 21 (23/88, 26%) and identified that aneuploidy of chromosomes 9 (5/88, 6%) and 17 (7/88, 8%) were recurrent in both groups. From comparison of patients with IGH-CRLF2 and P2RY8-CRLF2, we noted a higher frequency of IKZF1 (79% v 37% p<0.001) and BTG1 (50% v 8%, p<0.001) deletions and greater complexity (>3 additional copy number alterations (CNA)) by MLPA (SALSA MLPA P335 kit, MRC Holland, The Netherlands) (p=0.037) among IGH-CRLF2 patients. We verified that (48/160) 30% of CRLF2 -d patients were Down syndrome (DS).In addition to CNA within the immunoglobulin loci and the genes investigated by MLPA, SNP6.0 arrays (Affymetrix, Santa Clara, USA) (n=25) and subsequent FISH screening identified recurrent CNA in the histone cluster at 6p22.2 (7/25, 28%), VPREB1 (6/25, 24%), ADD3 (14/58, 24%), BTLA (4/25, 16%), SLX4IP (12/43, 28%), SERP2 and TSC22D1 (6/53, 11%), PBX3 (16/54, 30%). A novel finding was an interstitial deletion at Xp11.4 fusing USP9X to DDX3X (9/50, 18%).A total of 137 somatic structural variants (SV) were detected by whole genome sequencing of 11 patients [tandem duplication- (n=6), deletion- (n=104), inversion- (n=14) and translocation-type (n=13)] creating an average of 12.7 validated SV per patient. Mutations were identified at an incidence of 12.8 per patient, including mutations in JAK1 and JAK2 (40%) and IL7R, CACNA1D and USH2A (n=2). Mutations in kinase genes were common and unique to all patients, a number have not been previously reported (e.g. ERBB4, TTBK1 and STK38L). Of interest, 7 patients had ≥1 point mutation in gene(s) involved in cell adhesion (e.g. NPNT, ADAM29, ITGB7, COL3A1 and USH2A).There was a significant difference in the median age and white blood cell count (WBC) between P2RY8-CRLF2 and IGH-CRLF2 (4yrs v 14yrs, p<0.001 and WBC >50x106 /L - 25% v 44%, p=0.021). Overall IGH-CRLF2 patients had a worse outcome compared to P2RY8-CRLF2 (p=0.191) with outcome heterogeneity in relation to gender (IGH females, HR 3.75, p=0.003), number of CNA (IGH <3 genes deleted, HR 3.79, p=0.047) and treatment trial (IGH have worse outcome on ALL97, HR 3.15, p=0.014) being observed. There was no heterogeneity for age (p=0.458), WBC (p=0.331), cytogenetic risk (p=1.0), MRD (p=1.0) or DS-ALL (p=0.877).In summary we have confirmed differences in age and WBC, involvement of kinase gene mutations and chromosomal gains in CRLF2-d ALL. Other findings included recurrent deletions involving TSC22D1 and PBX3, transcription factors involved in other human cancers; SLX4IP, an uncharacterized gene, frequently deleted in childhood ALL; the novel involvement of USP9X, a deubiquitinase involved in cancer development. We also identified recurrent mutations in cell adhesion genes. We showed that IGH-CRLF2 patients had a worse outcome, with heterogeneity seen between gender, number of gene deletions and treatment trial. We have shown that there are genetic and clinical differences dependent on whether CRLF2 is deregulated by IGH or P2RY8, suggesting that they should be regarded as different clincial entities in future studies. DisclosuresCampbell:14M genomics: Other: Co-founder and consultant.

Abstract A12: Genomically unstable glioblastoma are sensitive to Parp inhibition

Abstract Background: The development of effective targeted drugs for the treatment of glioblastoma (GBM) represents a major unmet need. Veliparib (ABT-888; Abbvie) inhibits both PARP1 and PARP2 (poly[ADP-ribose] polymerase). The successful clinical application of veliparib and other PARP inhibitors (PARPi) will be assisted by the identification of predictive biomarkers. Materials and Methods: We performed whole genome sequencing (WGS) on ten GBM specimens with matched normal DNA. Mutations were detected using qSNP and GATK and indels called with Pindel and GATK. Somatic structural variants were identified using the qSV package. For all ten specimens we developed patient derived cell lines (PDCLs). We tested the efficacy of the combination of veliparib and temozolomide (TMZ) on all of the PDCLs. Results: Strikingly, in 2 patient samples denoted as G89 and G54, the mutation rate was at least 16-fold higher compared to the other 8 participants. G89 and G54 exhibited a large number of single nucleotide variants (SNVs), insertions, deletions and structural variant (SV) events (&amp;gt;2 million) when compared to the other 8 patient samples. The scale of genomic instability suggested defects in DNA maintenance, which could potentially define sensitivity to DNA damaging agents. Upon closer examination of the mutational profile of G89 and G54, mutations were found in at least one of the mismatch repair (MMR) genes: MLH1, MSH2, MSH6 and PMS2. We also detected mutations in other genes involved in DNA maintenance such as XRCC4, FANCA, FANCD2, ATR, RPA1, REV3L and PARP1. We designated these patients' GBM as U-GBM class. We found the U-GBM class of tumors, G89 and G54 to be hypersensitive to the combination of TMZ/veliparib. Conclusion: Mutations in DNA maintenance pathways may be a method for selecting patients for therapies involving the combination of DNA damaging agents such as radiotherapy, and PARP inhibitors. Additionally, the signature associated with genomically unstable GBM may be a method of identifying potential responders to PARP inhibitor therapy. Citation Format: Kerrie L. McDonald, Mustafa Khasraw, Toni Rose Jue, Swapna Joshi, Julia Yin. Genomically unstable glioblastoma are sensitive to Parp inhibition. [abstract]. In: Proceedings of the AACR Special Conference: Advances in Brain Cancer Research; May 27-30, 2015; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2015;75(23 Suppl):Abstract nr A12.

ATPS-54GENOMICALLY UNSTABLE GLIOBLASTOMA (U-GBM) SHOW EXQUISITE SENSITIVITY TO PARP INHIBITION

BACKGROUND: The development of effective targeted drugs for the treatment of glioblastoma (GBM) represents a major unmet need. Veliparib (ABT-888; Abbvie) inhibits both PARP1 and PARP2 (poly[ADP-ribose] polymerase). The successful clinical application of veliparib and other PARP inhibitors (PARPi) will be assisted by the identification of predictive biomarkers. METHODS: We performed whole genome sequencing (WGS) on ten GBM specimens with matched normal DNA. Mutations were detected using qSNP and GATK and indels called with Pindel and GATK. Somatic structural variants were identified using the qSV package. We tested the efficacy of the combination of veliparib and radiotherapy (RT) in vitro using patient derived cell lines (PDCLs) and in vivo using our patient derived xenograft (PDX) model. RESULTS: Strikingly, in 2 patient samples denoted as G89 and G54, the mutation rate was at least 16-fold higher compared to the other 8 participants. G89 and G54 exhibited a large number of single nucleotide variants (SNVs), insertions, deletions and structural variant (SV) events (>2 million) when compared to the other 8 patient samples. The scale of genomic instability suggested defects in DNA maintenance, which could potentially define sensitivity to DNA damaging agents. Upon closer examination of the mutational profile of G89 and G54, mutations were found in at least one of the mismatch repair (MMR) genes: MLH1, MSH2, MSH6 and PMS2. We also detected mutations in other genes involved in DNA maintenance such as XRCC4, FANCA, FANCD2, ATR, RPA1, and REV3L. We designated these patients' GBM as U-GBM class. We found the U-GBM class of tumors, G89 and G54 to be hypersensitive to the combination of veliparib and RT, both in vitro and in vivo. CONCLUSION: Mutations in DNA maintenance pathways may be a method for selecting patients for therapies involving the combination of DNA damaging agents such as radiotherapy, and PARP inhibitors.

"PEAR-ing" Genomic and Epigenomic Analyses for Cancer Gene Discovery.

Somatic structural variants in tumor genomes can deregulate transcription through repositioning of enhancer elements. A new method, PEAR-ChIP, leverages paired-end H3K27ac chromatin immunoprecipitation combined with high-throughput sequencing and current computational methods to identify such events.

Whole-Genome Sequencing of Growth Hormone (GH)-Secreting Pituitary Adenomas.

The somatic landscape of pituitary adenomas is largely unknown. Identification of somatic alterations aims at better understanding of tumor pathology. The objective of the study was a genome-wide characterization of somatic single-nucleotide variants, structural variants, and copy-number aberrations in somatotropinomas. Whole-genome sequencing and single-nucleotide polymorphism array analyses were performed on 12 fresh-frozen somatotropinomas and their corresponding blood samples. All the coding somatic variants were confirmed by Sanger sequencing. Studied tumors were somatotropinomas. Apart from one AIP mutation-positive patient, all cases were mutation negative for the established germline mutations associated with pituitary adenomas. There were no interventions. Somatic variants were identified with an established computational pipeline and filtered against germline data. Somatic copy number alteration analyses were performed using segmentation-based approaches. A genome-wide analysis revealed on average 129 somatic single-nucleotide variants per tumor. Further analysis of coding regions showed on average 2.3 single-nucleotide variants per tumor. The only recurrent somatic events were the oncogenic GNAS mutation (p.Arg201Cys) and shared chromosome losses (chromosomes 1, 6, 13, 14, 15, 16, 18, 22). Analysis of somatic structural variants revealed one tumor with a complex chromosomal rearrangement. Somatotropinomas showed a low number of somatic genetic alterations. Whereas no novel recurrently mutated genes could be identified, the somatic landscape has potential to affect the Ca(2+) and ATP pathways known to be involved in the pituitary tumorigenesis. Further studies, eg, methylome and transcriptome analyses, are needed to investigate possible interplay between the recurrent chromosome losses and epigenetic factors.

Local sequence assembly reveals a high-resolution profile of somatic structural variations in 97 cancer genomes.

Genomic structural variations (SVs) are pervasive in many types of cancers. Characterizing their underlying mechanisms and potential molecular consequences is crucial for understanding the basic biology of tumorigenesis. Here, we engineered a local assembly-based algorithm (laSV) that detects SVs with high accuracy from paired-end high-throughput genomic sequencing data and pinpoints their breakpoints at single base-pair resolution. By applying laSV to 97 tumor-normal paired genomic sequencing datasets across six cancer types produced by The Cancer Genome Atlas Research Network, we discovered that non-allelic homologous recombination is the primary mechanism for generating somatic SVs in acute myeloid leukemia. This finding contrasts with results for the other five types of solid tumors, in which non-homologous end joining and microhomology end joining are the predominant mechanisms. We also found that the genes recursively mutated by single nucleotide alterations differed from the genes recursively mutated by SVs, suggesting that these two types of genetic alterations play different roles during cancer progression. We further characterized how the gene structures of the oncogene JAK1 and the tumor suppressors KDM6A and RB1 are affected by somatic SVs and discussed the potential functional implications of intergenic SVs.

SpeedSeq: ultra-fast personal genome analysis and interpretation.

SpeedSeq is an open-source genome analysis platform that accomplishes alignment, variant detection and functional annotation of a 50× human genome in 13 hours on a low-cost server, alleviating a bioinformatics bottleneck that typically demands weeks of computation with extensive hands-on expert involvement. SpeedSeq offers competitive or superior performance to current methods for detecting germline and somatic single nucleotide variants, indels, and structural variants, and includes novel functionality for streamlined interpretation.

Abstract 5245: Identification and quantification of somatic structural variations in tumor, urine and plasma from bladder cancer patients

Abstract Bladder tumors can be divided into two clinically divergent groups, non-muscle invasive (NMIBC) and muscle invasive bladder tumors (MIBC). NMIBCs have a recurrence rate of 50-70%, while MIBCs often develop metastases. No molecular biomarkers are yet available to directly guide treatment and follow-up regimes for patients with BC. Aim of this study was to identify tumor specific, large somatic structural variations (SSV) in DNA from tumors and quantify SSVs in DNA from urine and in circulating free DNA (cfDNA) in plasma from successive visits during treatment and follow-up. Methods: Whole genome paired-end (WGS) or mate-pair (MPS) sequencing on HiSeq2000 and Nextseq sequencers. SSVs called by algorithms “CREST” or “Breakdancer”. cfDNA was extracted from 1-3 ml plasma. Quantification of SSVs was done by droplet digital PCR (ddPCR) using probes covering the breakpoint. We selected six patients with progressive/metastatic disease, followed for 4 to 19 years with consecutive sampling of tumors, urine and plasma samples (4 to 27 visits). WGS was carried out on DNA from germ-line and tumors from four patients (57x and 45x read depth). CREST identified 20-226 SSVs per tumor. PCR assays (n = 170, 21-64 per patient) showed validation rates of 60-75% (CTX), 75-90% (ITX), 88-100% (INS) and 50% (INV) among the eight tumors. MPS was carried out on DNA from two patients (22-30x physical read depth). Breakdancer identified 82 and 120 SSVs, 24 and 13 were potential somatic with a validation rate of 69% and 79%. For each patient, 3-5 fluorescent probes covering the breakpoint were used to quantify the level of tumor specific by ddPCR. SSVs were quantified in tumor DNA from several visits. Their frequency varied considerably (0.08-0.54) between the individual tumors, indicating a large inter-tumor heterogeneity. SSVs detected in DNA from urine pellets were correlated to the presence of tumor and treatment for some cases. Inter-individual cfDNA level ranged from 600-3500 genomic equivalents (GEs) per ml plasma. SSVs were identified in plasma. The level of ctDNA ranged between 5-300 GEs/ml plasma and an increase in ctDNA was already detectable up to 21 months before progression to muscle invasive disease. For metastasized disease, a 42-fold increase of the ctDNA level was observed compared to localized disease. No SSVs were detected in cfDNA from patients cured from BC (3 and 7 years post-recurrence and 5 years post-cystectomy). Conclusions: Tumor derived SSVs were identified and quantified in cfDNA from plasma of patients with progressive and metastatic BC. Detection of SSVs in cfDNA was possible at early time points before progression and metastasis of the disease were discovered by cystoscopy or image diagnostic. The tumor specific and highly sensitive detection and quantification of SSVs in cfDNA from plasma may offer a unique tool to monitor disease course, residual tumor burden, disease progression and treatment response in BC. Citation Format: Karin Birkenkamp-Demtröder, Iver Nordentoft, Emil Christensen, Søren Hoyer, Thomas Reinert, Søren Vang, Jørgen Bjerggaard Jensen, Torben F. Ørntoft, Lars Dyrskjøt. Identification and quantification of somatic structural variations in tumor, urine and plasma from bladder cancer patients. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 5245. doi:10.1158/1538-7445.AM2015-5245

Abstract 475: Identification of SHANK2 as a tumor suppressor disrupted by recurrent somatic structural variation (SV) in neuroblastoma

Abstract Background. Neuroblastoma is a malignancy of the developing sympathetic nervous system which exacts significant morbidity and mortality in children. Large sequencing efforts have revealed a relative paucity of somatic point mutations in this and other childhood cancers. Structural variations (SVs) including translocations, inversions, deletions, duplications, and other complex events can occur in germline DNA or be acquired somatically in tumors. Recurrent somatic SVs affecting known genes are likely to be functional and may elucidate novel tumor suppressors or oncogenes. Methods. We are performing whole-genome sequencing (WGS) of 187 matched tumor-normal pairs from high-risk patients as part of the NCI-TARGET project (http://www.ocg.cancer.gov/programs/target) using Complete Genomics technology. High confidence somatic SVs were identified and a subset subsequently validated with Sanger sequencing; in vitro functional studies were performed using human derived neuroblastoma cell line models. Results. To date, we have analyzed WGS data from 106 matched stage 4 tumor-normal pairs. Among 2,770 high-confidence somatic SVs (average 26.1 per tumor, range 1-158), we observed recurrent focal deletions in ZFHX3 (n = 5) and EZH2 (n = 2), in addition to ATRX (n = 2), ARID1B (n = 2), and PTPRD (n = 5), each of the latter genes previously reported in neuroblastoma. A total of 210 inter-chromosomal translocations disrupting known genes were detected (average 2.0 per tumor, range 0-12). Notably, 14.4% of MYCN non-amplified stage 4 tumors (11/76) harbored SV breakpoints within SHANK2 at 11q13; all SHANK2 breakpoints were validated by Sanger sequencing. With the exception of a single deletion at the c-terminus, all breakpoints mapped within the long isoform of SHANK2 (transcript variant 1: NM_012309), and clustered within the ankyrin repeat domain which is involved in protein-protein interactions. Low SHANK2 expression in primary tumors obtained at diagnosis was associated with worse overall survival (p = 6.3 × 10−5), suggesting SHANK2, known to regulate neuronal differentiation, may function as a tumor suppressor. Consistent with this hypothesis, mRNA expression of SHANK2 was low in a panel of 27 neuroblastoma cell lines and further silencing with siRNA did not produce an observable phenotype. In contrast, forced expression of the long isoform of SHANK2 in three independent neuroblastoma cell lines with low endogenous SHANK2 levels resulted in profound reduction in cell growth (p &amp;lt; 0.0001) and viability (p &amp;lt; 0.0001) as measured by RTCES and CellTiter-glo assays respectively. Conclusion. A plethora of SVs exist in neuroblastoma, and unlike somatic point mutations, many are recurrent. Studies are ongoing to determine the mechanism by which SHANK2 suppresses neuroblastoma tumorigenesis and to understand the biological relevance of other recurrent coding and non-coding SVs in this childhood malignancy. Citation Format: Karina Conkrite, Nicole Ferraro, Lee McDaniel, Derek A. Oldridge, Edward Attiyeh, Shahab Asgharzadeh, Maura Diamond, Jaime Guidry Auvil, Tanja Davidsen, Malcom Smith, Wendy B. London, Robert Seeger, Javed Khan, Daniela S. Gerhard, John M. Maris, Sharon J. Diskin. Identification of SHANK2 as a tumor suppressor disrupted by recurrent somatic structural variation (SV) in neuroblastoma. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 475. doi:10.1158/1538-7445.AM2015-475

Abstract 4876: An integrated pipeline for detecting and characterizing structural variation in cancer

Abstract Detection and characterization of somatic structural variants (SVs) and copy-number variants (CNVs) from whole genome sequencing remains a challenging part of cancer analysis. Many callers have been developed that use different detection strategies, but most methods suffer from high rates of false positives and false negatives, and agreement between different callers is usually low. We have developed a flexible pipeline that combines the results of multiple callers, filters calls to remove likely artifacts, and functionally annotates the resulting variants. We employ a diverse set of variant callers utilizing a combination of read depth, read pair, and split read detection methods: NBIC-seq (Xi et al., 2011), Crest (Wang et al., 2011), Delly (Rausch et al., 2012), and BreakDancer (Chen et al., 2009). To remove artifact calls due to mis-mapping, we apply filters that discard predicted SVs whose breakpoints exhibit certain sequence features (e.g. extensive mapping ambiguity, high repeat content). SVs corresponding to known germline variants (1000G, DGV, in-house database) are marked and removed as unlikely somatic variants: this greatly helps to prevent both sequencing protocol- and caller-specific artifacts as well as false positive somatic calls arising from missed calls in the matched germline sample. Finally, we employ our sensitive split read mapper SplazerS to identify SV breakpoints with base pair precision. In this step, we are also able to remove remaining germline variants for which we find split read support in the matched normal sample. The final predicted structural variants are annotated for overlap with SVs in COSMIC, overlap with known cancer genes and potential impact on gene structure. We use a public synthetic data set (DREAM challenge; Boutros et al., 2014) to demonstrate that using our selected ensemble of tools significantly improves sensitivity as compared to any single caller and that our filters effectively remove artifacts. Further, we show results from a set of colorectal cancer samples (Brannon et al., 2014) in which highly similar primary and metastatic tumors show excellent agreement in somatic SV calls in the absence of overlap between unrelated samples. Results from testing our pipeline on TCGA glioblastoma multiforme tumors, for which validated genomic rearrangements are available, will also be presented. In conclusion, our pipeline improves detection of SVs by integrating orthogonal calling methods and facilitates identification of clinically relevant SVs through effective filters and cancer-specific functional annotation. Citation Format: Minita Shah, Dayna M. Oschwald, Soren Germer, Michael C. Zody, Toby Bloom, Anne-Katrin Emde. An integrated pipeline for detecting and characterizing structural variation in cancer. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4876. doi:10.1158/1538-7445.AM2015-4876

Abstract 2176: Joint structural variant analysis of colorectal cancer whole genome sequencing data

Abstract Colorectal cancer (CRC) is one of the most common cancers, with over million cases and over 600,000 deaths each year in the world. It is estimated that a significant part of CRCs may have a hereditary compo- nent, but majority of this heritability remains unexplained. Structural variants (SVs) comprise an important type of variation that, until the availability of deep-coverage whole-genome sequencing (WGS), has been hard to study in genome-wide scale. It has been suggested that a significant portion of unexplained, or “missing”, heritability could be explained by structural variants. Somatic structural aberrations are known to play a role in tumorigenesis and tumor progression. However, structural variation discovery is often complicated by low sequencing coverage, sequencing artifacts, repeat regions and errors of the reference genome. To study the role of structural variation in CRC, we analyzed deep-coverage WGS data of matched tumor-normal sample pairs of 213 Finnish CRC patients and additional in-house WGS data. In these data, we identified both germline and somatic structural variants using a recent structural variant caller DELLY. A novel algorithm was used to combine DELLY paired-end structural variant calls across normal and tumor samples to improve both sensitivity and specificity of variant detection. A total of 4,854,576 unfiltered structural events were identified, consisting of 2,912,552 deletions, 1,304,636 tandem duplications, 71,724 inversions and 565,664 interchromosomal translocations. To call high-confidence somatic events, we identified 475,902 structural events in tumors that were not present in any normal sample. Calling accuracy was further improved by annotating events by mappability and proximity to repeat regions. Our results demonstrate the applicability of integrating structural variant data across a large number of samples to obtain high-quality structural variant calls. Citation Format: Esa Pitkänen, Tatiana Cajuso, Riku Katainen, Sofie Lundgren, Sari Tuupanen, Outi Kilpivaara, Lauri A. Aaltonen. Joint structural variant analysis of colorectal cancer whole genome sequencing data. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2176. doi:10.1158/1538-7445.AM2015-2176

Abstract 4742: Using 1ng of DNA to detect haplotype phasing and gene fusions from whole exome sequencing of cancer cell lines

Abstract We have used a new platform from 10X Genomics to obtain ultra-deep and long-range exome sequencing data from 1ng of cancer cell line DNA. While traditional Whole Exome Sequencing (WES) methods have inputs of ∼50-500ng and lack long-range data, we demonstrate that the 10X platform can successfully obtain both gene fusion and haplotype phasing information along with standard SNP and indel variant-calling from WES on an Illumina HiSeq using just 1ng of DNA. The 10X platform massively partitions long DNA into &amp;gt;100,000 individual reactions. Each partition generates sequencing libraries with a unique barcode; the integrated downstream analysis platform uses the barcodes to map Illumina's short reads back to original long DNA fragments. Because the barcoding is performed prior to exome enrichment, the long-range information conveyed by the barcodes is retained even after standard hybrid capture enrichment. Powerful algorithms are implemented using a custom informatics toolset that is self-contained and easy to install, and the resulting structural variant calls can be easily visualized and explored using 10X's haplotype-aware genome browser. We used the 10X platform to analyze cancer cell lines with previously reported gene fusions. In all cases described, the fusion breakpoints were in introns of &amp;gt; 5.9kbp, and would therefore not be detectable with standard WES. In lung cancer cell line NCI-H2228, we called two annotated fusions, EML4/ALK, and ALK/PTPN3. Two fusion genes (MCHR2/ANO4, and CR627240/SACS) were identified in glioblastoma cell line U-87MG, and at least four (CD44/FGFR2, PVT1/PDHX, PVT1/ATE1, PVT1/PPAPDC1A) in gastric cancer cell line SNU-16. Long-range information provided by the technology also enables haplotype phasing of exome samples. Our algorithms phased the vast majority of SNPs within genes, with phase blocks up to 1.25Mbp. We used this haplotype information to phase somatic structural variants, including large-scale deletions and amplifications. To achieve the sensitivity required to call somatic mutations from low purity tumor specimens, it is necessary to be able to attain high coverage depth. This can be a significant challenge for low inputs; however we demonstrate duplicate-removed depth of &amp;gt;200x from inputs as low as 1ng, and sensitivity and precision for variant calling from 1ng libraries that is equivalent to that from ligation-based libraries prepared from 200ng of DNA. It is becoming increasingly apparent that a complete understanding of tumor genomic profiles requires the ability to detect structural variation and phasing information from low input and high coverage sequencing libraries. With this study we demonstrate the ability of the 10X platform to enable such applications from ultra-deep targeted sequencing data using only 1ng of DNA. Citation Format: Mirna Jarosz, Michael Schnall-Levin, Grace X. Y. Zheng, Patrick Marks, Sofia Kyriazopoulou-Panagiotopoulou, Patrice Mudivarti, Kristina Giorda. Using 1ng of DNA to detect haplotype phasing and gene fusions from whole exome sequencing of cancer cell lines. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4742. doi:10.1158/1538-7445.AM2015-4742

Abstract 4871: A new multiple feature approach for rapid and highly accurate somatic structural variation discovery from whole cancer genome sequencing

Abstract We present SWAN, a statistical framework with implemented software package for accurate detection of genomic structural variants in next-generation sequencing (NGS) data from cancer. SWAN is designed to sensitively detect submicroscopic and lesser size genome structural variants that could possibly present at low allele frequency in the sample as may be the case in clonal subpopulations of a primary tumor. The likelihood ratio scan statistics module of SWAN collectively quantifies evidence from read-pair, hang-reads based on Marked Poisson Process models while the softclipping module additionally integrates the evidence from remapped soft-clipped reads. Using a spike-in dataset, we compared SWAN against established structural variant callers in recovering simulated homozygous and heterogeneous variants of various sizes and types. SWAN maintained high accuracy, high precision and low false positives across this broad spectrum and outperformed all the other callers. In additional benchmark with golden standard real data from sample NA12878, SWAN again showed the best-balanced accuracy among all callers and was 30% more sensitivity in detecting validated deletions than the next runner-up caller LUMPY in 50x coverage. Even at 5x coverage, SWAN was able to find 40% of the validated deletions while LUMPY did less than 20%. As an application to whole cancer genome sequencing, we analyzed a matched sequenced colorectal adenocarcinoma sample and found 39 somatic deletions, insertions and duplications, some of which could potentially drive the cancer progression. SWAN is quite efficient in memory utilization and speed, and can be used for large-scale genome-wide structural variant genotyping and discovery based on NGS data. Citation Format: Li C. Xia, John Bell, Jiamin Chen, Nancy R. Zhang, Hanlee P. Ji. A new multiple feature approach for rapid and highly accurate somatic structural variation discovery from whole cancer genome sequencing. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 4871. doi:10.1158/1538-7445.AM2015-4871

Somatic structural variation and cancer.

With the advent of next-generation sequencing technologies, we have witnessed a rapid pace of discovery of new patterns of somatic structural variation in cancer genomes, and an attempt to figure out their underlying mechanisms. Some of these mechanisms are associated with particular cancer types, and in some cases are the main cause of the structural mutations that drive the oncogenic process. This review provides an overview of the patterns of somatic structural variation and chromosomal structures that characterize cancer genomes, their causal mechanisms and their impact in oncogenesis.

Structural variation discovery in the cancer genome using next generation sequencing: computational solutions and perspectives.

Somatic Structural Variations (SVs) are a complex collection of chromosomal mutations that could directly contribute to carcinogenesis. Next Generation Sequencing (NGS) technology has emerged as the primary means of interrogating the SVs of the cancer genome in recent investigations. Sophisticated computational methods are required to accurately identify the SV events and delineate their breakpoints from the massive amounts of reads generated by a NGS experiment. In this review, we provide an overview of current analytic tools used for SV detection in NGS-based cancer studies. We summarize the features of common SV groups and the primary types of NGS signatures that can be used in SV detection methods. We discuss the principles and key similarities and differences of existing computational programs and comment on unresolved issues related to this research field. The aim of this article is to provide a practical guide of relevant concepts, computational methods, software tools and important factors for analyzing and interpreting NGS data for the detection of SVs in the cancer genome.

Analysis of circulating tumour DNA to monitor disease burden following colorectal cancer surgery

ObjectiveTo develop an affordable and robust pipeline for selection of patient-specific somatic structural variants (SSVs) being informative about radicality of the primary resection, response to adjuvant therapy, incipient recurrence and...

GE-23 * ENHANCER HIJACKING ACTIVATES GFI1 FAMILY ONCOGENES IN MEDULLOBLASTOMA

Medulloblastoma is a highly malignant paediatric brain tumour currently treated with a combination of surgery, radiation and chemotherapy, posing a considerable burden of toxicity to the developing child. Genomics has illuminated the extensive intertumoural heterogeneity of medulloblastoma, identifying four distinct molecular subgroups. Group 3 and Group 4 subgroup medulloblastomas account for most paediatric cases; yet, oncogenic drivers for these subtypes remain largely unidentified. In an effort to uncover novel drivers in these poorly characterized subgroups, we analyzed whole-genome sequencing data derived from 137 primary Group 3 and Group 4 medulloblastoma samples and performed a systematic, high-resolution screen for chromosomal breakpoints recurrently targeting putative candidate genes. This comprehensive analysis of structural variants identified highly disparate genomic structural variants, restricted to Groups 3 and 4, resulting in specific and mutually exclusive activation of the growth factor independent 1 family proto-oncogenes, GFI1 and GFI1B. Diverse mechanisms of structural variation, including duplications, deletions, inversions, translocations, and other complex variants were observed in nearly all GFI1/GFI1B-activated cases. Integration of genome sequencing data with sample-matched ChIP-seq data for enhancer histone marks established that somatic structural variants juxtapose GFI1 or GFI1B coding sequences proximal to active enhancer elements, including super-enhancers, instigating their oncogenic activity. Functional experiments performed in mice confirmed the oncogenicity of both GFI1 and GFI1B in the context of medulloblastoma and demonstrated apparent functional synergy between both of these candidates and MYC. These studies establish GFI1 and GFI1B as novel, highly prevalent medulloblastoma oncogenes specifically active in Group 3 and Group 4. Given their high frequencies of activation, GFI1 and GFI1B represent excellent candidates for prioritization of molecularly targeted therapy aimed at treatment of a significant proportion of Group 3 and Group 4 medulloblastoma patients. Moreover, our data implicates ‘enhancer hijacking’ as an efficient mechanism capable of driving oncogene activation in a childhood cancer.

Comprehensive characterization of complex structural variations in cancer by directly comparing genome sequence reads.

The development of high-throughput sequencing technologies has advanced our understanding of cancer. However, characterizing somatic structural variants in tumor genomes is still challenging because current strategies depend on the initial alignment of reads to a reference genome. Here, we describe SMUFIN (somatic mutation finder), a single program that directly compares sequence reads from normal and tumor genomes to accurately identify and characterize a range of somatic sequence variation, from single-nucleotide variants (SNV) to large structural variants at base pair resolution. Performance tests on modeled tumor genomes showed average sensitivity of 92% and 74% for SNVs and structural variants, with specificities of 95% and 91%, respectively. Analyses of aggressive forms of solid and hematological tumors revealed that SMUFIN identifies breakpoints associated with chromothripsis and chromoplexy with high specificity. SMUFIN provides an integrated solution for the accurate, fast and comprehensive characterization of somatic sequence variation in cancer.

Abstract 2370: Comparative evaluation of somatic mutations calls on single nucleotide variants and structural variants using breast cancer cell lines

Abstract In an effort to improve somatic mutation calling pipelines, the Human Genome Sequencing Center at Baylor College of Medicine in collaboration with The Cancer Genome Atlas has undertaken a comprehensive sequencing and comparative analysis of mutation calling in the context of paired tumor/normal samples. Comparative evaluation of somatic mutation calling of single nucleotide variants (SNVs) and structural variants (SVs) has been performed using two tumor/normal paired breast cancer cell lines, HCC1954 and HCC1143. Through controlled in silico mixing of the tumor/normal and alternating cell line sequencing data, the effects of tumor purity (i.e. normal contamination) and subclonal expansion on the sensitivity and specificity of somatic mutation calling has been comprehensively evaluated. In addition, cross sequencing-platform comparisons have been assessed, evaluating the cost/benefits of including long range information for structural variation analysis; for example, PacBio long-read or large-fragment mate pair sequencing libraries, and RNAseq gene expression information for association of functional consequences due to somatic mutation. A full spectrum of sequencing data has been generated and undergone SNV and SV somatic mutation calling and integrative analysis; including: whole genome sequencing (WGS) from Illumina short paired-end (250X-coverge), Illumina long mate-pair (10X-coverage), and PacBio long read (5X-coverage). For the short paired-end WGS data, combinations of various fractions of tumor-derived and normal-derived cell line DNA sequence were used to simulate tumor purity. Similarly, additions of varying fractions of known spike-in SNVs and SVs were used to simulate subclonal expansions on a heterogeneous tumor genome background. This benchmarking study provides novel insights on the benefits of incorporating long-range sequencing data; such as: Illumina 3Kbp mate-pair and PacBio long read (9Kbp N50 read-length), allowing for orthogonal sequencing library and cross-platform confirmation, as well as independent identification of somatic SV mutations. Furthermore, incorporation of tumor/normal RNAseq data allows integration of gene expression and fusion transcripts thereby associating functional consequences with specific somatic mutations. Lastly, comparison of copy number alterations from paired-end WGS read-depth and single nucleotide polymorphism (SNP) array data were are comparatively evaluated. This comprehensive comparative evaluation and integrative analysis of somatic SNV and SV calling across a full spectrum of sequencing data provides essential feedback that is useful for improving somatic mutation calling pipelines and aiding in the design of future whole genome sequencing cancer studies. Citation Format: Oliver A. Hampton, Richard A. Gibbs, David A. Wheeler. Comparative evaluation of somatic mutations calls on single nucleotide variants and structural variants using breast cancer cell lines. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 2370. doi:10.1158/1538-7445.AM2014-2370