Abstract Whole-chromosome and segmental copy-number changes are nearly ubiquitous in human cancers. Here, we present the first pan-cancer landscape of haplotype-specific somatic copy number alterations (SCNAs) in nearly 9,000 cancers across 30 cancer types from The Cancer Genome Atlas (TCGA) whole-genome sequencing (WGS) data. This analysis was primarily carried out with CancerVision, a proprietary bioinformatic workflow for detecting both somatic and germline variants in cancer samples developed by Inocras. The haplotype-specific copy-number analysis provides three pieces of information not available from previous analysis. First, the haplotype-specific SCNAs enables a more accurate assessment of aneuploidy, i.e., the fraction of the cancer genome with copy number alterations. Second, the haplotype resolution directly resolves interactions between germline variant genotypes and SCNAs. Finally, haplotype-specific SCNAs directly inform the instigating mechanisms of chromosomal instability. In this study, we provide examples demonstrating each scenario. Notably, we observed a diverse range of haplotype-specific SCNA patterns, each reflecting a distinct mechanism of chromosomal instability, including arm-level alterations related to WGD, segmental changes indicative of breakage-fusion-bridge (BFB), complex rearrangements displaying signatures of successive BFB cycles and chromothripsis, as well as focal amplifications consistent with extrachromosomal DNA (ecDNA) formation. Our analysis of haplotype-specific SCNAs in TCGA suggests a mechanism-based classification of copy-number patterns linked to chromosomal instability. Extending this framework to treatment-exposed samples could provide new insights into synthetic lethality dependencies, with potential diagnostic and therapeutic implications for cancer precision medicine. Citation Format: Chunyang Bao, Matthew Leventhal, Hansol Park, Gang-Hee Lee, Ryul Kim, Won-Chul Lee, Jonghoon Lee, Yoonsuh Lee, Beomki Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P. Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Cheng-Zhong Zhang, Young Seok Ju. Genomic signatures of chromosomal instability from a pan-cancer landscape of haplotype-specific copy-number alterations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 5934.

Abstract Background: The initial TCGA exome sequencing project established a foundational catalog of somatic mutations. However, capture biases and limited coverage in GC-rich and repetitive regions may have obscured bona fide driver events and introduced systematic “blind spots” in the landscape of protein-coding genes. Leveraging high-depth, PCR-free whole-genome sequencing (WGS), we revisited the somatic mutation landscape in multiple tumor types to enhance driver discovery by improving sensitivity in GC-rich regions and reducing artifact-driven false positives. Methods: We analyzed matched tumor–normal PCR-free WGS data from >8,000 TCGA cases across several cancer types with existing previously published somatic mutation calls from whole exome sequencing (WES) data, applying a unified best-practice pipeline for SNV/indel detection and stringent post-calling filters. We focused on (i) concordance between WGS and exome across coding regions, (ii) coverage, allele fraction, and local sequence context (GC content, segmental duplications) of discordant sites, and (iii) recurrence and positional clustering of variants in established cancer genes. Results: PCR-free WGS identified substantially more high-confidence coding mutations than WES, with the greatest gains in GC-rich exons and difficult-to-capture loci. This increased sensitivity uncovered additional pathogenic or likely pathogenic variants in canonical drivers, including TP53 and FOXA1 in breast cancer, EGFR in glioblastoma, and BAP1 in uveal melanoma (UVM), thereby strengthening known genotype–phenotype associations. In UVM, we observed an indel in the BAP1 5′UTR/promoter region that was systematically missed by WES but supported by robust read evidence in WGS data, nominating an expanded spectrum of BAP1-disrupting events with potential regulatory and clinical relevance. Conversely, we found a subset of WES-only calls localized to segmental duplications or low-support sites, consistent with technical artifacts. Integrated re-annotation of discordant calls thus both eliminates spurious events and reveals underappreciated driver mutations and mechanisms. Conclusions: Our systematic comparison demonstrates that PCR-free WGS can refine the somatic mutation and driver landscape beyond exome-based catalogs, particularly in GC-rich and regulatory regions. These results advocate for re-interrogation of previously profiled tumor types with contemporary WGS to achieve a more complete and accurate map of cancer-driving alterations. Citation Format: Gengchao Wang, David I Heiman, Vasuki Narasimha Swamy, Chip Stewart, Xavi Loinaz, Ron Solan, Chunyang Bao, David Lehotzky, Brian P Danysh, Luis Antonio Corchete Sanchez, Zachary Everton, Sam Wiseman, Antonia Kowalewski, Samantha Van Seters, Saveliy Belkin, Haruna Tomono, Andrew D Cherniack, Ryul Kim, Gang-Hee Lee, Won-Chul Lee, Hansol Park, Rebecca Jang, Young Seok Ju, Gad Getz, Esther Rheinbay. Pan-cancer PCR-free whole-genome sequencing refines the somatic driver landscape beyond exome sequencing alone [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1990.

Abstract Overexpression of LINE-1 (L1) elements is recognized as a hallmark of many cancers, yet their activity and the number of somatic LINE-1 retrotranspositions (soL1Rs) vary markedly across tumor types. Despite their ubiquity in certain cancers, such as esophageal and colorectal carcinomas, the biological significance of transposable elements in cancer development remains underappreciated. To elucidate the contribution of L1 activity to oncogenesis, we analyzed whole-genome sequencing (WGS) data from >8,000 tumor-normal pairs in The Cancer Genome Atlas (TCGA). In total, we identified over 132,000 soL1R events, with their distribution differing widely across cancer types. The highest soL1R burdens were observed in esophageal carcinoma (ESCA, 134.2 events per case), bladder carcinoma (BLCA, 74.0), head and neck squamous cell carcinoma (HNSC, 56.3), lung squamous cell carcinoma (LUSC, 53.3), colon adenocarcinoma (COAD, 29.7), and rectal adenocarcinoma (READ, 16.8). The number of transductions correlated strongly with the overall soL1R burden in each cohort. Activity of germline LINE-1 source elements, inferred from transduction counts, identified 22q12.1-2 as the most active locus, followed by Xp22.2-2, 5q14.3-2, and 14q23.1. Using L1 transduction events, we further discovered 12 previously unreported germline LINE-1 source loci. Functional analysis revealed that soL1R events contribute to oncogenesis through both direct disruption of tumor suppressor genes and complex genomic rearrangements mediated by retrotransposition. In summary, this population-scale WGS analysis delineates the pan-cancer landscape of somatic LINE-1 retrotranspositions, uncovers novel germline source elements, and highlights the multifaceted role of soL1R activity in cancer genome evolution. Citation Format: Beomki Lee, Ryul Kim, Chunyang Bao, Hansol Park, Gang-Hee Lee, Jonghoon Lee, Yoonsuh Lee, Won-Chul Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P. Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Cheng-Zhong Zhang, Young Seok Ju. Pan-cancer LINE-1 retrotransposition landscapes in TCGA whole-genome sequences [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1980.

Abstract Recent genomic foundation models have advanced DNA sequence interpretation, yet most remain constrained to local sequence patterns and fail to produce the patient-level insights required for clinical decision-making. To address this limitation, we developed a framework that extends beyond sequence-level inference, enabling robust patient stratification through a Cancer Foundation Model. Our approach begins with “DNAChunker”, which employs a dynamic H-net-based tokenization strategy that divides the genome into variable-length segments, preserving high-resolution detail in regulatory and coding regions while efficiently compressing repetitive sequences. When evaluated on the Nucleotide Transformer and Genomic Benchmarks, DNAChunker achieved performance comparable to the state-of-the-art GENERator (1.2 billion parameters) while using only 156 million parameters. To translate these genomic embeddings into patient-level insights, we implemented a transformer-based Cancer Aggregation Model that integrates mutation embeddings with somatic copy-number alteration (SCNA) features. The framework was evaluated on large whole-genome sequencing (WGS) cohorts, including PCAWG (n=2,040) and CUBRICS breast cancer samples (n=1,053), with TCGA-BRCA (breast cancer; n=920) serving as an external validation cohort. The model effectively stratified patients by cancer type (accuracy, 96.89%), homologous recombination deficiency (HRD; accuracy, 92.83%), and PAM50 subtype (accuracy, 84.05%). Notably, it classified PAM50 intrinsic subtypes using only DNA-level information, eliminating the conventional reliance on RNA-based expression profiling. The Cancer Foundation Model demonstrates that patient-level representation learning from whole-genome data can achieve clinically meaningful stratification across diverse tumor types. By bridging the gap between genomic sequence interpretation and actionable phenotypic classification, this framework establishes a foundation for AI-based precision oncology. With further validation, it will facilitate biomarker discovery and patient stratification in clinical trials directly from WGS data. Citation Format: Jonghoon Lee, Chunyang Bao, Hansol Park, Gang-Hee Lee, Yoonsuh Lee, Beomki Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Young Seok Ju, Won-Chul Lee, Ryul Kim. AI-Driven stratification of cancer patients using The Cancer Genome Atlas whole-genome sequencing data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7268.

Abstract Background: Despite the well-documented male bias in the incidence and mortality of most non-sex-specific cancers, the Y chromosome (chrY) remains as a genomic blind spot. Major pan-cancer genomic consortia, including The Cancer Genome Atlas (TCGA) and PCAWG, have historically excluded the chrY from somatic analyses due to computational challenges related to its haploid state, repetitive structure, and homology with the X chromosome. Consequently, the somatic landscape of chrY alterations, including complete loss of Y (LOY), structural variants (SVs), single-nucleotide variants (SNVs), and INDELs, remains largely uncharacterized. Methods: We implemented a pan-cancer computational framework utilizing deep whole-genome sequencing (WGS) data from a comprehensive re-analysis of 26 cancer types from the TCGA cohort. We developed and applied a novel bioinformatic pipeline to accurately call somatic LOY, complex SVs, SNVs, and INDELs. The pipeline was specifically designed to overcome analytical challenges of the chrY, including its repetitive structure and the correct handling of pseudoautosomal regions (PARs). Potential artifacts were flagged using variant-support metrics and by identifying breakpoints in problematic regions. Results: The analysis revealed a highly heterogeneous landscape of chrY alterations across tumor types. We found a high prevalence of somatic SVs in some tumor types, such as bladder (BLCA), head and neck (HSNC), stomach (STAD), and prostate (PRAD) cancers, while others, like thyroid (THCA) and renal clear cell carcinoma (KIRC), were rarely affected. In affected tumors, the SV landscape included numerous inter-chromosomal translocations, which often showed low variant allele frequency (VAF), suggesting a subclonal origin. In contrast, the clonality of chrY deletions varied across cancer types. The SNV/INDEL burden was substantial in certain cancers, for example, in glioblastoma (GBM) and PRAD, but was concentrated in non-coding regions, with protein-coding mutations being infrequent. Interestingly, we observed a chrY SNV spectrum enriched in C>T and T>C substitutions across several tumor types. Conclusion: We developed a computational framework to systematically analyze the Y chromosome across multiple cancer types, overcoming historical analytical challenges. Our results reveal a diverse spectrum of somatic alterations, including structural variants and point mutations. These findings provide the first comprehensive, pan-cancer map of the somatic chrY landscape. Citation Format: Luis Antonio Corchete Sanchez, Chunyang Bao, Saveliy Belkin, Andrew D. Cherniack, Brian P. Danysh, Zachary Everton, David I. Heiman, Ryul Kim, Antonia Kowalewski, Gang-Hee Lee, Won-Chul Lee, David Lehotzky, Xavi Loinaz, Vasuki Narasimha Swamy, Hansol Park, Ron Solan, Chip Stewart, Haruna Tomono, Samantha Van Seters, Gengchao Wang, Sam Wiseman, Young Seok Ju, Gad Getz, Esther Rheinbay. The pan-cancer landscape of chromosome Y alterations [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1986.

Abstract Mutations not only drive carcinogenesis but also record the evolutionary history of cancer. Homologous recombination deficiency (HRD), a defect in high-fidelity DNA double-strand break repair, produces characteristic mutational footprints that define an HRD phenotype. HRD has major clinical relevance as a predictor of response to DNA-damaging agents and targeted therapies, yet current HRD assays were developed primarily in breast and ovarian cancers, and their applicability to other tumor types remains unclear. We conducted a pan-cancer whole-genome analysis of >8,000 primary tumors from The Cancer Genome Atlas (TCGA), spanning ∼30 cancer types, including ∼900 breast and 300 ovarian cancers. We developed HRDecide, an HRDetect-derived framework that uses comprehensively annotated germline and somatic mutations in homologous recombination (HR) pathway genes, including structural variants and non-canonical splice alterations informed by recent advances in variant-effect prediction, to refine HRD identification across cancer types. Applying HRDecide to TCGA whole-genome sequencing data, we identified HRD-positive tumors and delineated their mutational signatures. HRD-associated features were broadly shared across cancers, extending beyond breast and ovarian tumors to include tumor types such as prostate, pancreatic, gastric, and hepatocellular carcinomas. These signatures encompassed canonical short deletions with microhomology as well as cancer type-specific patterns across substitution, indel, and structural variation profiles. Notably, we also characterized atypical HRD-like phenotypes that showed only a subset of components of the classical HRD signatures. The spectrum of inactivated HR pathway genes, and the predominant modes of disruption, point mutation, structural variation, and allelic loss, varied substantially by cancer type. The scale of this cohort enabled robust associations between HRD phenotypes and pathogenic mutations in HR pathway genes, such as ATM, CHEK2, and RAD51B, often truncated by structural variations and LINE-1 retrotranspositions, extending beyond the classical BRCA1/2, PALB2, and RAD51C events reported in PCAWG and Hartwig datasets. In summary, this pan-cancer study provides a comprehensive whole-genome landscape of HRD in primary tumors, revealing tissue-specific modes of repair failure and expanding the HRD spectrum beyond the BRCA-centered paradigm. These findings establish a reference framework for tissue-agnostic HRD biomarker development and therapeutic stratification across cancer types. Citation Format: Joonoh Lim, Chunyang Bao, Hansol Park, Gang-Hee Lee, Ryul Kim, Won-Chul Lee, Jonghoon Lee, Yoonsuh Lee, Beomki Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P. Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Young Seok Ju. Whole-genome sequencing of >8,000 TCGA samples reveals diverse and atypical spectra of homologous recombination deficiency in primary human cancers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1977.

In estrogen receptor-positive (ER +) breast cancer, CDK4/6 inhibitors (CDK4/6is) combined with endocrine therapy (ET) are standard first-line treatment for metastatic disease. However, most patients eventually develop resistance. Activating ESR1 mutations are a prevalent mechanism of acquired resistance to ET and are enriched after ET plus a CDK4/6 inhibitor (CDK4/6i), but their role in the clonal evolution and adaptive mechanisms of acquired resistance to CDK4/6 inhibition, independent of ET, is unknown. In addition, whether different CDK4/6is impose distinct selective pressures and divergent resistance states remains elusive. To investigate the clonal dynamics, cell states and cellular plasticity during acquired CDK4/6i resistance in mutant versus wild-type (WT) ESR1, we performed high-complexity DNA barcoding (ClonTracer library) with longitudinal sampling and multi-omic profiling in an isogeneic MCF7 model expressing WT ER or Y537S mutant ER. We also evaluated the clonality of the ESR1 mutations in clinical samples with CDK4/6i resistance. We showed that ESR1 mutations are enriched in clinical tumors with acquired resistance to CDK4/6is, and in paired biopsies expanded to near clonality after treatment. We demonstrated progressive clonal selection with both divergent and partially convergent evolutionary trajectories. The ESR1 mutation substantially reshapes clonal and epigenetic evolution during palbociclib resistance but had a weaker impact under abemaciclib selection. Overall, clonal evolution and cell states in palbociclib and abemaciclib resistance were distinct. Single-cell RNA-seq revealed transcriptional heterogeneity highlighting cellular plasticity during passaging of cells and selection. Finally, in vivo barcoding of mammary xenograft, local recurrences, and distant metastases demonstrated site-specific clonal outgrowth in mutant ER metastases, and partial overlap between metastatic and CDK4/6i-resistant subclones, supporting the dual role of specific populations in therapeutic resistance and metastatic colonization. High-resolution lineage tracing and multi-omic studies demonstrate that CDK4/6i resistance is shaped by clonal selection and adaptive remodeling of cell states, with the ESR1 mutation status and the specific inhibitor acting as key determinants of evolutionary trajectories. These findings suggest that both variables should be considered when designing sequential and combination treatment strategies to overcome CDK4/6i resistance.

Abstract Cancer driver genes are oncogenes and tumor suppressor genes whose changes in function or expression promotes tumorigenesis. They are used to study cancer behavior, to classify cancer types, and to guide precision medicine. Cancer driver genes are usually identified by statistical analysis of coding mutations in tumor genomes. However, studies searching for new drivers are limited by statistical power, since mutations in each specific driver are often rare, and drivers are often specific to tumor subtypes. In addition, studies relying on whole-exome sequencing miss functionally important noncoding regions, such as promoters, and sequencing using PCR amplification often misses GC-rich and GC-poor regions. As a result, PCR-free whole-genome sequencing (WGS) of larger cohorts is required to continue the discovery of cancer drivers. To comprehensively identify single nucleotide variant (SNV) and indel drivers in cancer, we analyzed >8,000 tumor-normal pairs spanning 31 cancer types from The Cancer Genome Atlas (TCGA) sequenced using PCR-free WGS. We identified somatic SNVs and indels using a custom pipeline, then used MutSig2CV and dNdScv to identify genes under positive selection in coding regions and used Dig to detect increased mutation rates in coding and non-coding regions, including 5'-UTR, 3'-UTR, and promoter sequences. Our analysis identified well-known noncoding drivers, including TERT promoter mutations, TP53 5’-UTR mutations, and NFKBIZ 3’-UTR mutations, as well as some novel candidate cancer drivers, including in non-coding regions. We believe that the new drivers we discovered will provide new insights into the genetic mechanisms of tumorigenesis, aid in the development of cancer therapeutics, and offer a foundation for the use of noncoding driver events in precision oncology. Citation Format: David Lehotzky, Ron Solan, Antonia Kowalewski, Nick Haradhvala, Xavier Loinaz, Hansol Park, Vasuki N. Swamy, David Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Chunyang Bao, Andrew Cherniack, Luis A. Corchete Sanchez, Brian P. Danysh, Zachary Everton, Ryul Kim, Gang-Hee Lee, Won-Chul Lee, Chip Stewart, Haruna Tomono, Gengchao Wang, Young Seok Ju, Esther Rheinbay, Gad Getz. Discovery of coding and non-coding driver mutations across >8,000 TCGA whole genomes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2001.

Abstract Copy number alterations are frequent, early events in cancer evolution. Previous pan-cancer analyses have not characterized recurrent copy number alterations on the actively and inactively transcribed X chromosome. The monoallelic expression of the active X chromosome predicts a higher sensitivity of cell fitness to copy number alterations on the active X chromosome than expected from autosomal copy number alterations. Large deletions on the active X chromosome are likely intolerable due to loss of expression of essential genes; a duplication of the active X chromosome will increase chromosome X expression by 100% instead of 50%,which is the expected increase of gene expression from duplication of a single autosome homolog. Recurrent copy number alterations on the X chromosomes could lead to the discovery of new, therapeutically actionable drivers of tumor evolution. We have developed a new method to accurately determine the copy number alterations on the active and inactive X from combined whole genome sequencing and RNA-seq of >8,000 TCGA tumors. We found recurrent copy number gains on the active X chromosome q-arm spanning the telomere. The active X chromosome copy number gains overlapped previously identified cancer genes in the Cancer Gene Census from other tumors such as ATP2B3, FLNA, MTCP1, RPL10, PHF6 and GPC3. These segmental and arm-level gains were not significantly more frequent in patients with whole genome doubling or TP53 mutations. These results suggested that recurrent gains of the q-arm of the active X chromosome were not a result of genomic instability, suggesting that these gains were putative driver events that could be truncal in tumor evolution. Telomere-spanning segmental copy number gains on the q-arm were less common in female glioblastoma or uveal melanoma tumors, which are cancer types that affect both male and female patients. This work shows that copy number gains on the active X chromosome are a new source of cancer driver mutations. The sensitivity of the active X chromosome to selection suggests that genes in recurrent X chromosome amplifications could be new genetic dependencies or therapeutic targets in treating female cancers. The results of this study may identify previously unknown drivers of cancer on the active X chromosome that could be new, effective therapeutic targets in cancer. Citation Format: Matthew Joseph Leventhal, Chunyang Bao, Ron Solan, Haruna Tomono, Andrew D. Cherniack, Luis A. Corchete Sanchez, Rebecca Jang, Jeehee Suh, Antonia Kowalewski, Sam Wiseman, Samantha Van Seters, Saveliy Belkin, David I. Heiman, Chip Stewart, David Lehotzky, Vasuki N. Swamy, Brian P. Danysh, Gengchao Wang, Xavi Loinaz, Zachary Everton, Gang-Hee Lee, Won-Chul Lee, Hansol Park, Ryul Kim, Young Seok Ju, Esther Rheinbay, Gad Getz, Cheng-Zhong Zhang. The landscape of X chromosome copy number alterations in cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 2002.

Abstract Somatic copy number alterations (SCNAs) are frequent oncogenic events. Previous studies have identified recurrent gene-level SCNAs using DNA microarray and whole exome sequencing data. These studies have also identified focal amplifications of regulatory regions. High-throughput whole genome sequencing enables deeper, comprehensive analysis of these focal SCNAs in non-coding regions, potentially revealing novel drivers of oncogenesis. To this end, we systematically analyzed SCNAs and structural variants in WGS data from >8,000 tumor-normal pairs across 31 cancer types from The Cancer Genome Atlas (TCGA) using GISTIC2.0 to identify focal, recurrent promoter and enhancer SCNAs near known oncogenes and tumor suppressor genes. We leveraged structural variant calls to orthogonally validate our primary focal SCNA findings. We identified a recurrent amplification event specific to the EGFR promoter/enhancer region in a subset of IDH-wildtype glioblastoma samples. This suggests an additional SCNA-driven mechanism for EGFR upregulation, distinct from canonical gene body amplification or point mutations, in a fraction of these aggressive tumors. Our ongoing work aims to broaden the landscape of regulatory SCNAs across cancers and validate their transcriptional impact to provide new biological insights and candidate therapeutic targets. Citation Format: Haruna Tomono, Chunyang Bao, Antonia Kowalewski, David Lehotzky, Ron Solan, Matthew Leventhal, Luis Antonio Corchete Sanchez, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Brian P. Danysh, Chip Stewart, Vasuki Narasimha Swamy, Gengchao Wang, Xavi Loinaz, Zachary Everton, Gang-Hee Lee, Won-Chul Lee, Hansol Park, Ryul Kim, Young Seok Ju, Esther Rheinbay, Gad Getz, Andrew D. Cherniack, Matthew L. Meyerson, Rameen Beroukhim. Copy number analysis of regulatory regions reveal recurrent promoter/enhancer somatic copy number alterations across >8,000 TCGA samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1978.

Abstract Comprehensively modeling tumor evolution is important for cancer diagnosis, treatment, and minimal residual disease (MRD) monitoring. Liquid biopsies enable non-invasive sampling of tumor DNA during a patient’s cancer treatment. Accurate detection of low allele fraction somatic variants in circulating tumor DNA (ctDNA) has clinical importance for cancer detection and monitoring. Current ctDNA panel sequencing methods lack the breadth of variants captured in whole genome sequencing (WGS) and cannot fully track tumor clones over time. Moreover, deep and accurate cell-free DNA (cfDNA) sequencing is challenging due to constraints of low tumor fraction in the blood and the high cost of WGS sequencing. Ultima Genomics (UG) developed paired plus-minus sequencing (ppmSeq) technology, a whole-genome duplex sequencing approach that lowers observed error rates and allows for a high yield of duplex molecules. Using UG’s deep WGS ppmSeq, we developed our mutation calling algorithms to enable sensitive detection of somatic variants in cfDNA samples, allowing us to track clonal populations over time. To test whether we could identify shared clonal populations in patient sequences between other platforms and ppmSeq technology, we collected post-mortem tissue specimens from 12 patients of various cancer types (breast, cholangiocarcinoma, etc.) and 38 pre-mortem cfDNA samples from the same patients. First, these patients’ data were sequenced with existing Illumina sequencing technologies (WES) and then on the UG sequencing platform for both tissue and ppmSeq. We sequenced the ppmSeq data to an average depth of 135x (range 104x to 208x) and achieved an average duplex proportion of 38% (range 31% to 44%). For all patients’ tissue and cfDNA samples that were sequenced with both ppmSeq WGS (UG) and WES (Illumina), we reconstructed phylogenetic trees to determine the life history of the cancer using our PhylogicNDT suite of tools. Phylogenies reconstructed from ppmSeq WGS cfDNA identified many more mutations and richer trees than those reconstructed from WES of the same cfDNA samples. We compared the cfDNA ppmSeq phylogenies with those constructed from WGS of the same patient’s tissue samples and identified the same clonal populations, enabling us to match clones from the blood to the tissue and track the progression of individual clones. The identification of large clonal populations in ctDNA sequenced with Ultima Genomics’ ppmSeq approach is an improvement over methods with lower sensitivity and coverage; moreover, this approach can identify clonal populations that change over a given treatment course. This more accurate inference of cancer evolution information could enable us to better guide therapies and identify novel mechanisms of resistance that would have been missed with prior methods. Citation Format: Elizabeth E. Martin, Julian Hess, Carrie Cibulskis, Mendy Miller, Brian P. Danysh, Chip Stewart, Elena Helman, Ilya Soifer, Doga C. Gulhan, Dejan Juric, Doron Lipson, Gad Getz. Leveraging Ultima Genomics ppmSeq WGS-cfDNA to accurately detect clonal evolution over sequential blood biopsies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3992.

Abstract The systematic identification of therapeutically actionable genomic alterations across tumor types is essential to advance precision oncology. Using the CancerVisionTM whole-genome analysis platform, we analyzed >8,000 whole-genome sequencing (WGS) samples spanning more than 30 cancer types to characterize clinically actionable mutations. Actionability was defined according to the Cancer Knowledgebase evidence levels, where Level A denotes biomarkers linked to FDA-approved on-label therapies. Overall, 2903 patients (∼32%) harbored at least one Level A actionable alteration encompassing ∼20,100 events (median 2 per sample). Level A on-label targets were detected in more than half of THCA (thyroid cancer), KIRC (kidney clear-cell), SKCM (skin cutaneous melanoma), BRCA (breast cancer), and COAD (colon adenocarcinoma), underscoring their clinical relevance. The ten most frequent Level A targets were PIK3CA, KRAS, BRAF, VHL, PTEN, ERBB2, BRCA1/2, NRAS, and EGFR, showing variable frequencies among cancer types. Among KRAS alterations, p.G12C was the predominant actionable hotspot, mainly in lung adenocarcinoma with a few cases in colorectal and rectal cancers. A total of 950 fusion targets were identified across the cohort. BRCA exhibited the highest frequency (45 cases; 2.5%), including 40 ESR1-CCDC170 fusions, followed by THCA (55 cases; 9.7%), dominated by CCDC6-RET (25 cases). NRG1 fusions were found in 145 samples (1.7%), markedly higher than the historical pan-cancer frequency (∼0.2%). Other recurrent targetable fusions included NTRK (0.7%; 60 samples), RET (0.7%), ALK (0.7%), BRAF (0.6%), and FGFR (2.5%; 210 samples), the latter largely involving FGFR2 (∼180 cases). Across all fusion classes, >80% retained the kinase domain, supporting oncogenic potential. Clinical-trial-matched targets were most commonly TP53, PIK3CA, and CDKN2A, reflecting their broad inclusion in precision-medicine studies. This comprehensive pan-cancer analysis defines the most extensive WGS-based landscape to date of actionable mutations and druggable fusions. The whole-genome approach enabled high-resolution detection of rare structural variants that are often missed by targeted sequencing panels, highlighting the potential of whole-genome profiling to uncover therapeutically relevant but under-recognized targets, supporting the integration of WGS into future tumor-agnostic clinical trial design. Citation Format: Ryul Kim, Chunyang Bao, Hansol Park, Gang-Hee Lee, Won-Chul Lee, Jonghoon Lee, Yoonsuh Lee, Beomki Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Young Seok Ju. Uncovering therapeutically targetable mutations from The Cancer Genome Atlas (TCGA) whole-genome datasets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 492.

Abstract Structural variants (SVs)—large-scale genomic deletions, duplications, inversions, and translocations—can promote tumorigenesis by activating proto-oncogenes, disrupting tumor suppressors, generating oncogenic fusions, rewiring gene regulation, and mediating catastrophic events such as chromoplexy and chromothripsis. Yet, the role of SVs as cancer-driving mutations remains less comprehensively characterized than that of single-nucleotide variants or indels, largely due to the historic scarcity of tumor whole-genome sequencing data required for accurate SV detection. In this study, we analyzed whole-genome sequencing data from >8,000 tumor-normal pairs spanning 31 cancer types from The Cancer Genome Atlas (TCGA) to systematically characterize SVs at unprecedented scale. Compared with the flagship Pan-Cancer Analysis of Whole Genomes (PCAWG) project, our analysis includes roughly four times as many samples and six additional cancer types. Somatic SVs were identified using a custom pipeline combining Manta and dRanger with optimized downstream filters. Across all tumors, we detected >1 million somatic SVs. We developed two complementary frameworks to interpret these variants. First, to classify SVs by their genomic architecture, we inferred genomic segments and their associated copy number alterations driven by SVs ranging from a single to hundreds of breakpoints, enabling the generalization of distinct patterns through unsupervised clustering. Second, to identify candidate driver genes, we developed SVelfie, a statistical framework that detects genes significantly enriched in functional SVs predicted to confer gain- or loss-of-function effects. Applying SVelfie to 385 prostate and 333 ovarian cancer genomes revealed multiple novel candidate driver genes, with additional discoveries expected as analysis extends to the full >8,000-sample dataset. This work represents the most comprehensive analysis to date of SV drivers in cancer. By uniting large-scale WGS data with new computational frameworks for SV classification and driver detection, we expand the catalog of SV-driven cancer genes, illuminate mechanisms of SV-mediated oncogenesis, and advance the clinical utility of whole-genome sequencing in precision oncology. Citation Format: Antonia Kowalewski, Xavi Loinaz, Hansol Park, Vasuki Narasimha Swamy, David Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Chunyang Bao, Andrew D. Cherniack, Luis A. Corchete Sanchez, Brian P. Danysh, Zachary Everton, Ryul Kim, Gang-Hee Lee, Won-Chul Lee, David Lehotzky, Ron Solan, Chip Stewart, Haruna Tomono, Gengchao Wang, Rameen Beroukhim, Young Seok Ju, Esther Rheinbay, Gad Getz. Systematic discovery and classification of structural variant drivers across >8,000 TCGA whole genomes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1989.

Abstract The goal of The Cancer Genome Atlas (TCGA) has been to continually characterize the genomic and transcriptomic landscapes across diverse malignancies. In this analysis, we assess matched tumor-normal Whole-Genome Sequencing (WGS) data from a previously sequenced set of adult cancer patients to identify germline pathogenic variations. These latest TCGA data consist of cancer patients from varied ancestral backgrounds with solid tumors and lymphoid cancers and patient-matched normal samples, consisting of blood or tissue. We have now obtained high-quality tumor and normal WGS data from over 8,000 samples, including normal samples derived mainly from the patient-matched blood samples. To discover novel links between genetics and cancer predisposition, we are analyzing germline variants—including single-nucleotide variants (SNVs) and structural variants (SVs)—in known cancer genes, as well as in genes with recurrent mutations that may have been missed or previously could not be assessed using exome or low-coverage genome studies.We have conducted a preliminary analysis focused on identifying Pathogenic or Likely Pathogenic (P/LP) variants, classified according to American College of Medical Genetics and Genomics guidelines, in cancer predisposition genes. This initial pass on a majority of samples confirmed the presence of P/LP variants across the cohort in canonical cancer predisposition genes, including BRCA1, BRCA2, ATM, and other genes integral to the mismatch and DNA repair pathways. Across cohorts, we identified known P/LP germline variants in established cancer predisposition genes in less than 10% of all cases, with the largest number of variants identified in BRCA1/2, aligning with previous published work. The prevalence of P/LP variants was not uniform across cancer types, with some showing an enrichment, including breast cancer, while other cancer types, like low-grade gliomas, demonstrated a lower prevalence than the average. These findings underscore the highly variable, tumor-specific landscape of germline predisposition.We continue to leverage this comprehensive WGS dataset as a key resource for ongoing analyses of germline-somatic interactions. We are actively investigating how these germline P/LP variants shape the tumor's somatic mutational landscape and contribute to oncogenesis. We are also investigating whether haplotype-specific copy number alterations contribute to the pathogenicity of identified germline cancer risk alleles. Further work includes characterizing both structural and complex non-coding pathogenic variants that were previously inaccessible in exome or low-coverage WGS studies. Citation Format: Ryul Kim, Owen Hirschi, Matthew Leventhal, Chunyang Bao, Hansol Park, Gang-Hee Lee, Won-Chul Lee, Jonghoon Lee, Yoonsuh Lee, Beomki Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Cheng-Zhong Zhang, Matthew L. Meyerson, Young Seok Ju. Germline predisposition in The Cancer Genome Atlas (TCGA) whole-genome sequencing datasets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1991.

Abstract Cancer arises from the progressive accumulation of genomic alterations. The Cancer Genome Atlas (TCGA), a landmark consortium project, has comprehensively characterized 33 cancer types through multi-omics profiling of over 11,000 tumor-normal pairs. However, most TCGA-based studies had relied on whole-exome sequencing (WES), which covers only ∼1-2% of the genome, leaving the majority of the genomic landscape unexplored. To achieve a more comprehensive understanding of cancer genomes, the Broad Institute and Inocras collaboratively analyzed TCGA whole-genome sequencing (WGS) data encompassing over 8,000 tumors across more than 30 cancer types, which were initially analyzed by whole-exome sequencing. To fully leverage this resource, we applied CancerVision, an automated and streamlined bioinformatics pipeline developed by Inocras for clinical-grade WGS interpretation. CancerVision detects diverse genomic variants, including single-nucleotide variants (SNVs), insertions/deletions (indels), somatic copy number alterations (SCNAs), structural variants (SVs), and germline mutations, while also inferring homologous recombination deficiency (HRD) and mutational signatures. Using CancerVision, we performed a comprehensive, harmonized reanalysis of the TCGA WGS dataset and benchmarked our results against the bioinformatics pipelines from the Broad Institute and the official TCGA exome data. Across representative cohorts, ovarian cancer (CNV-driven), thyroid cancer (SNV-driven), and glioblastoma (mixed), CancerVision achieved high concordance, often uncovering additional high-confidence genomic alterations not captured in the existing TCGA resource. By integrating these results, we expand the known landscape of somatic variants, improve driver gene detection, and demonstrate the power of whole-genome-based analytics for actionable insights in precision oncology. Citation Format: Chunyang Bao, Hansol Park, Gang-Hee Lee, Ryul Kim, Won-Chul Lee, Jonghoon Lee, Yoonsuh Lee, Beomki Lee, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Vasuki Narasimha Swamy, David I. Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Andrew D. Cherniack, Luis Antonio Corchete Sanchez, Brian P. Danysh, Zachary Everton, Chip Stewart, Haruna Tomono, Gengchao Wang, Esther Rheinbay, Gad Getz, Young Seok Ju. Comprehensive mutation profiling from The Cancer Genome Atlas (TCGA) whole-genome sequencing datasets [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7270.

Abstract Double-strand break repair leaves recognizable footprints in the genome. Among the most specific are short sequences inserted at structural-variant (SV) junctions—templated insertions often attributed to polymerase-θ-mediated end joining (TMEJ). Yet common readouts based on exact string matches overlook sequence background, distance from the break, and the topological context of candidate templates, inflating false positives and blurring mechanistic interpretation. We present a scalable statistical framework that infers templated insertions with controlled specificity by modeling alignment-score distributions against a distance-adjusted, genome-wide empirical null. The approach explicitly accommodates imperfect copying and partitions candidate templates into four breakpoint-proximal configurations, capturing positional and strand relationships that are informative of mechanisms. Applied to large somatic and germline whole-genome cohorts, the method reveals that template usage spans all configurations but differs systematically across cellular contexts. A notable fraction of events reflect imperfect copying, consistent with error-prone synthesis, whereas one configuration shows comparatively higher apparent fidelity—suggesting distinct biochemical routes within a broader TMEJ-like landscape. Configuration calls also stratify SV architecture: some are enriched in simple rearrangements while others localize to clustered, complex regions, indicating that local topology and repair pathway choice are linked. Beyond structure, configuration-specific burdens align with DNA-repair states and selected genotypes: contexts consistent with homologous-recombination deficiency show enrichment in particular configurations, while others display the opposite directionality, underscoring that “templated insertion” is not a single phenomenon but a family of related processes with diverging determinants. To enable cohort-scale analysis, we optimized the core alignment to produce full score matrices in a single pass and packaged the workflow into a containerized pipeline, yielding order-of-magnitude speedups and portable reproducibility. Together, these results establish a configuration-aware, statistically principled readout of templated insertions that is robust to sequence confounders and informative about mechanisms. Practically, the framework provides (i) a sharper lens for studying double-strand break repair in human samples, (ii) leads for repair-state biomarkers, and (iii) hypotheses connecting SV topology to polymerase usage. In doing so, it aims to move the field from anecdotal sequence sketches toward reproducible, cohort-scale inferences about double-strand break repair involving junctional insertions. Citation Format: Youyun Zheng, Gregory Raskind, Sophie Webster, Narmen Azazmeh, Haruna Tomono, Andrew Cherniack, David Lehotzky, Ron Solan, Antonia Kowalewski, Xavi Loinaz, Hansol Park, Vasuki N. Swamy, David Heiman, Samantha Van Seters, Saveliy Belkin, Sam Wiseman, Chunyang Bao, Luis A. Corchete Sanchez, Zachary Everton, Ryul Kim, Beomki Lee, Won-Chul Lee, Chip Stewart, Gengchao Wang, Brian P. Danysh, Young Seok Ju, Esther Rheinbay, Gad Getz, Rameen Beroukhim. Origins of structural variant junctional insertions across >8,000 TCGA whole genomes [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3247.

Clinical whole-exome sequencing (WES) has revolutionized clinical diagnostics by enabling scalable, cost-effective molecular profiling to detect somatic variants and identify novel therapeutic targets. In particular, the clinical evaluation of somatic variants relies on both high-quality sequencing data and robust variant detection with rapid turnaround times. We developed an updated WES workflow utilizing a co-developed Twist Bioscience targeting panel and the DRAGEN (Dynamic Read Analysis for GENomics) platform, benchmarked with cell line mixtures containing >40,000 simulated variants and a clinical cohort of FFPE tumor specimens. Our assay, the Broad Clinical Somatic Whole Exome Assay V6.0, demonstrated high sensitivity (96.9% for SNVs with variant allele fraction [VAF] >10% at ≥125X; 93.5% for InDels with VAF > 20% at ≥125X) and low false positive rates (0.04 and 0.01 per Mb, respectively). The assay meets clinical requirements and enables large-scale, accurate variant profiling of cancers, expanding opportunities for molecular diagnostics across diverse clinical settings.

Abstract Many cancers have inherent defects in DNA damage response (DDR) which influence their sensitivity to tumor-targeting therapies. Examples include increased activity of immunotherapies in cancers with mismatch-repair deficiency and sensitivity to PARP inhibitors and platinum-based therapies in the context of homologous recombination (HR) deficiency. However, we are currently unable to reliably determine which DDR defects are present in a given cancer sample, which severely limits our ability to exploit these therapeutic vulnerabilities. Structural variants (SVs), or genomic rearrangements formed as a product of aberrant double strand break repair, hold promise as biomarkers of DDR state. Indeed, SVs affect a larger proportion of the cancer genome than any other form of genetic alteration and have features that reflect their mechanism of formation. Here, we develop QuantHDP, a novel computational method for detecting SV signatures which leverages complex modeling of genetic features to distinguish between cancers with different DDR alterations and potentially identify clinically relevant biomarkers. QuantHDP models SV features with probability distributions that reflect our knowledge of the biological mechanisms that generate them, accounts for expected differences in signature content by cancer type, and automatically infers the number of signatures present in a given dataset. In addition to recapitulating multiple previously established associations with defects in DDR - including signatures of BRCA1, BRCA2, and CDK12 alterations — we also uncover novel signatures that warrant further investigation. Citation Format: Gregory Raskind, Youyun Zheng, Anthony Zhao, Julia Sun, Simona Dalin, Siyun Lee, Chunyang Bao, Antonia Kowalewski, Ron Solan, Sam Wiseman, Samantha Van Seters, Saveliy Belkin, David I. Heiman, Chip Stewart, David Lehotzky, Vasuki Narasimha Swamy, Brian P. Danysh, Luis Antonio Corchete Sanchez, Andrew D. Cherniack, Haruna Tomono, Gengchao Wang, Xavi Loinaz, Zachary Everton, Gang-Hee Lee, Won-Chul Lee, Hansol Park, Ryul Kim, Young Seok Ju, Gad Getz, Esther Rheinbay, Rameen Beroukhim. Structural variant signature discovery across >8,000 TCGA whole genomes using QuantHDP [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 7274.

Abstract With the release of high-quality resequencing of the TCGA cohorts, the scientific community has gained an opportunity to deepen its understanding of cancer’s underlying causes and create paths forward for its treatment. Well-experienced with these cohorts, we have developed (i) state-of-the-art computational pipelines that accurately characterize variants in sequenced cancer data, as well as (ii) the infrastructure to run these pipelines quickly and cost-effectively in the cloud. Using a combination of established tools and specialized filters that reduce the likelihood of false calls, our pipelines have been honed over years of use and run extensively in numerous cancer studies. The TCGA cohort presented a unique challenge of >8,000 deep coverage, PCR-free whole-genome-sequenced (WGS) tumor-normal pairs, requiring a massive upscaling of our pipeline infrastructure. Enhancements to individual tools and refinements of our cloud-based workflow engine have reduced the time and cost of our pipeline execution by more than 50% since the start of 2025, allowing us to characterize more than 8,000 of the pairs in less than 1 month. This undertaking detected more than 262 million mutations and 1.17 million structural variants that we believe will provide deep insight into cancer biology and potential therapeutic targets. Citation Format: Sam Wiseman, Samantha Van Seters, Saveliy Belkin, David I. Heiman, Vasuki Narasimha Swamy, Antonia Kowalewski, Scott Ritterbush, Zachary Everton, Ron Solan, Chip Stewart, David Lehotzky, Luis Antonio Corchete Sanchez, Xavi Loinaz, Haruna Tomono, Andrew D. Cherniack, Gengchao Wang, Brian P. Danysh, Young Seok Ju, Esther Rheinbay, Gad Getz. Fast and cost-effective cloud-based pipelines to analyze cancer sequencing data [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 6881.

Supplemental Figure 6: Evolutionary analysis of longitudinal plasma ctDNA samples