Pan-cancer genome and transcriptome analyses of 1,699 paediatric leukaemias and solid tumours.

Analysis of molecular aberrations across multiple cancer types, known as pan-cancer analysis, identifies commonalities and differences in key biological processes that are dysregulated in cancer cells from diverse lineages. Pan-cancer analyses have been performed for adult but not paediatric cancers, which commonly occur in developing mesodermic rather than adult epithelial tissues. Here we present a pan-cancer study of somatic alterations, including single nucleotide variants, small insertions or deletions, structural variations, copy number alterations, gene fusions and internal tandem duplications in 1,699 paediatric leukaemias and solid tumours across six histotypes, with whole-genome, whole-exome and transcriptome sequencing data processed under a uniform analytical framework. We report 142 driver genes in paediatric cancers, of which only 45% match those found in adult pan-cancer studies; copy number alterations and structural variants constituted the majority (62%) of events. Eleven genome-wide mutational signatures were identified, including one attributed to ultraviolet-light exposure in eight aneuploid leukaemias. Transcription of the mutant allele was detectable for 34% of protein-coding mutations, and 20% exhibited allele-specific expression. These data provide a comprehensive genomic architecture for paediatric cancers and emphasize the need for paediatric cancer-specific development of precision therapies.

Cancer germline predisposing variants and late mortality from subsequent malignant neoplasms among long-term childhood cancer survivors: a report from the St Jude Lifetime Cohort and the Childhood Cancer Survivor Study.

BACKGROUND: Carriers of cancer predisposing variants are at an increased risk of developing subsequent malignant neoplasms among those who have survived childhood cancer. We aimed to investigate whether cancer predisposing variants contribute to the risk of subsequent malignant neoplasm-related late mortality (5 years or more after diagnosis). METHODS: In this analysis, data were included from two retrospective cohort studies, St Jude Lifetime Cohort (SJLIFE) and the Childhood Cancer Survivor Study (CCSS), with prospective follow-up of patients who were alive for at least 5 years after diagnosis with childhood cancer (ie, long-term childhood cancer survivors) with corresponding germline whole genome or whole exome sequencing data. Cancer predisposing variants affecting 60 genes associated with well-established autosomal-dominant cancer-predisposition syndromes were characterised. Subsequent malignant neoplasms were graded using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.03 with modifications. Cause-specific late mortality was based on linkage with the US National Death Index and systematic cohort follow up. Fine-Gray subdistribution hazard models were used to estimate subsequent malignant neoplasm-related late mortality starting from the first biospecimen collection, treating non-subsequent malignant neoplasm-related deaths as a competing risk, adjusting for genetic ancestry, sex, age at diagnosis, and cancer treatment exposures. SJLIFE (NCT00760656) and CCSS (NCT01120353) are registered with ClinicalTrials.gov. FINDINGS: 12 469 (6172 male and 6297 female) participants were included, 4402 from the SJLIFE cohort (median follow-up time since collection of the first biospecimen 7·4 years [IQR 3·1-9·4]) and 8067 from the CCSS cohort (median follow-up time since collection of the first biospecimen 12·6 years [2·2-16·6]). 641 (5·1%) of 12 469 participants carried cancer predisposing variants (294 [6·7%] in the SJLIFE cohort and 347 [4·3%] in the CCSS cohort), which were significantly associated with an increased severity of subsequent malignant neoplasms (CTCAE grade ≥4 vs grade <4: odds ratio 2·15, 95% CI 1·18-4·19, p=0·0085). 263 (2·1%) subsequent malignant neoplasm-related deaths (44 [1·0%] in the SJLIFE cohort; and 219 [2·7%] in the CCSS cohort) and 426 (3·4%) other-cause deaths (103 [2·3%] in SJLIFE; and 323 [4·0%] in CCSS) occurred. Cumulative subsequent malignant neoplasm-related mortality at 10 years after the first biospecimen collection in carriers of cancer predisposing variants was 3·7% (95% CI 1·2-8·5) in SJLIFE and 6·9% (4·1-10·7) in CCSS versus 1·5% (1·0-2·1) in SJLIFE and 2·1% (1·7-2·5) in CCSS in non-carriers. Carrying a cancer predisposing variant was associated with an increased risk of subsequent malignant neoplasm-related mortality (SJLIFE: subdistribution hazard ratio 3·40 [95% CI 1·37-8·43]; p=0·0082; CCSS: 3·58 [2·27-5·63]; p<0·0001). INTERPRETATION: Identifying participants at increased risk of subsequent malignant neoplasms via genetic counselling and clinical genetic testing for cancer predisposing variants and implementing early personalised cancer surveillance and prevention strategies might reduce the substantial subsequent malignant neoplasm-related mortality burden. FUNDING: American Lebanese Syrian Associated Charities and US National Institutes of Health.

indelPost: harmonizing ambiguities in simple and complex indel alignments.

SUMMARY: Small insertions and deletions (indels) in nucleotide sequence may be represented differently between mapping algorithms and variant callers, or in the flanking sequence context. Representational ambiguity is especially profound for complex indels, complicating comparisons between multiple mappings and call sets. Complex indels may additionally suffer from incomplete allele representation, potentially leading to critical misannotation of variant effect. We present indelPost, a Python library that harmonizes these ambiguities for simple and complex indels via realignment and read-based phasing. We demonstrate that indelPost enables accurate analysis of ambiguous data and can derive the correct complex indel alleles from the simple indel predictions provided by standard small variant detectors, with improved performance over a specialized tool for complex indel analysis. AVAILABILITY AND IMPLEMENTATION: indelPost is freely available at: https://github.com/stjude/indelPost. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Pediatric Cancer Variant Pathogenicity Information Exchange (PeCanPIE): a cloud-based platform for curating and classifying germline variants.

Variant interpretation in the era of massively parallel sequencing is challenging. Although many resources and guidelines are available to assist with this task, few integrated end-to-end tools exist. Here, we present the Pediatric Cancer Variant Pathogenicity Information Exchange (PeCanPIE), a web- and cloud-based platform for annotation, identification, and classification of variations in known or putative disease genes. Starting from a set of variants in variant call format (VCF), variants are annotated, ranked by putative pathogenicity, and presented for formal classification using a decision-support interface based on published guidelines from the American College of Medical Genetics and Genomics (ACMG). The system can accept files containing millions of variants and handle single-nucleotide variants (SNVs), simple insertions/deletions (indels), multiple-nucleotide variants (MNVs), and complex substitutions. PeCanPIE has been applied to classify variant pathogenicity in cancer predisposition genes in two large-scale investigations involving >4000 pediatric cancer patients and serves as a repository for the expert-reviewed results. PeCanPIE was originally developed for pediatric cancer but can be easily extended for use for nonpediatric cancers and noncancer genetic diseases. Although PeCanPIE's web-based interface was designed to be accessible to non-bioinformaticians, its back-end pipelines may also be run independently on the cloud, facilitating direct integration and broader adoption. PeCanPIE is publicly available and free for research use.

Therapy-induced mutations drive the genomic landscape of relapsed acute lymphoblastic leukemia.

To study the mechanisms of relapse in acute lymphoblastic leukemia (ALL), we performed whole-genome sequencing of 103 diagnosis-relapse-germline trios and ultra-deep sequencing of 208 serial samples in 16 patients. Relapse-specific somatic alterations were enriched in 12 genes (NR3C1, NR3C2, TP53, NT5C2, FPGS, CREBBP, MSH2, MSH6, PMS2, WHSC1, PRPS1, and PRPS2) involved in drug response. Their prevalence was 17% in very early relapse (<9 months from diagnosis), 65% in early relapse (9-36 months), and 32% in late relapse (>36 months) groups. Convergent evolution, in which multiple subclones harbor mutations in the same drug resistance gene, was observed in 6 relapses and confirmed by single-cell sequencing in 1 case. Mathematical modeling and mutational signature analysis indicated that early relapse resistance acquisition was frequently a 2-step process in which a persistent clone survived initial therapy and later acquired bona fide resistance mutations during therapy. In contrast, very early relapses arose from preexisting resistant clone(s). Two novel relapse-specific mutational signatures, one of which was caused by thiopurine treatment based on in vitro drug exposure experiments, were identified in early and late relapses but were absent from 2540 pan-cancer diagnosis samples and 129 non-ALL relapses. The novel signatures were detected in 27% of relapsed ALLs and were responsible for 46% of acquired resistance mutations in NT5C2, PRPS1, NR3C1, and TP53. These results suggest that chemotherapy-induced drug resistance mutations facilitate a subset of pediatric ALL relapses.

RNAIndel: discovering somatic coding indels from tumor RNA-Seq data.

MOTIVATION: Reliable identification of expressed somatic insertions/deletions (indels) is an unmet need due to artifacts generated in PCR-based RNA-Seq library preparation and the lack of normal RNA-Seq data, presenting analytical challenges for discovery of somatic indels in tumor transcriptome. RESULTS: We present RNAIndel, a tool for predicting somatic, germline and artifact indels from tumor RNA-Seq data. RNAIndel leverages features derived from indel sequence context and biological effect in a machine-learning framework. Except for tumor samples with microsatellite instability, RNAIndel robustly predicts 88-100% of somatic indels in five diverse test datasets of pediatric and adult cancers, even recovering subclonal (VAF range 0.01-0.15) driver indels missed by targeted deep-sequencing, outperforming the current best-practice for RNA-Seq variant calling which had 57% sensitivity but with 14 times more false positives. AVAILABILITY AND IMPLEMENTATION: RNAIndel is freely available at https://github.com/stjude/RNAIndel. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

The landscape of coding RNA editing events in pediatric cancer.

BACKGROUND: RNA editing leads to post-transcriptional variation in protein sequences and has important biological implications. We sought to elucidate the landscape of RNA editing events across pediatric cancers. METHODS: Using RNA-Seq data mapped by a pipeline designed to minimize mapping ambiguity, we investigated RNA editing in 711 pediatric cancers from the St. Jude/Washington University Pediatric Cancer Genome Project focusing on coding variants which can potentially increase protein sequence diversity. We combined de novo detection using paired tumor DNA-RNA data with analysis of known RNA editing sites. RESULTS: We identified 722 unique RNA editing sites in coding regions across pediatric cancers, 70% of which were nonsynonymous recoding variants. Nearly all editing sites represented the canonical A-to-I (n = 706) or C-to-U sites (n = 14). RNA editing was enriched in brain tumors compared to other cancers, including editing of glutamate receptors and ion channels involved in neurotransmitter signaling. RNA editing profiles of each pediatric cancer subtype resembled those of the corresponding normal tissue profiled by the Genotype-Tissue Expression (GTEx) project. CONCLUSIONS: In this first comprehensive analysis of RNA editing events in pediatric cancer, we found that the RNA editing profile of each cancer subtype is similar to its normal tissue of origin. Tumor-specific RNA editing events were not identified indicating that successful immunotherapeutic targeting of RNA-edited peptides in pediatric cancer should rely on increased antigen presentation on tumor cells compared to normal but not on tumor-specific RNA editing per se.

Inhibition of SF3B1 by molecules targeting the spliceosome results in massive aberrant exon skipping.

The recent identification of compounds that interact with the spliceosome (sudemycins, spliceostatin A, and meayamycin) indicates that these molecules modulate aberrant splicing via SF3B1 inhibition. Through whole transcriptome sequencing, we have demonstrated that treatment of Rh18 cells with sudemycin leads to exon skipping as the predominant aberrant splicing event. This was also observed following reanalysis of published RNA-seq data sets derived from HeLa cells after spliceostatin A exposure. These results are in contrast to previous reports that indicate that intron retention was the major consequence of SF3B1 inhibition. Analysis of the exon junctions up-regulated by these small molecules indicated that these sequences were absent in annotated human genes, suggesting that aberrant splicing events yielded novel RNA transcripts. Interestingly, the length of preferred downstream exons was significantly longer than the skipped exons, although there was no difference between the lengths of introns flanking skipped exons. The reading frame of the aberrantly skipped exons maintained a ratio of 2:1:1, close to that of the cassette exons (3:1:1) present in naturally occurring isoforms, suggesting negative selection by the nonsense-mediated decay (NMD) machinery for out-of-frame transcripts. Accordingly, genes involved in NMD and RNAs encoding proteins involved in the splicing process were enriched in both data sets. Our findings, therefore, further elucidate the mechanisms by which SF3B1 inhibition modulates pre-mRNA splicing.

Estimated number of adult survivors of childhood cancer in United States with cancer-predisposing germline variants.

PURPOSE: To estimate the absolute number of adult survivors of childhood cancer in the U.S. population who carry a pathogenic or likely pathogenic variant in a cancer predisposition gene. METHODS: Using the Surveillance, Epidemiology, and End Results (SEER) Program, we estimated the number of childhood cancer survivors on December 31, 2016 for each childhood cancer diagnosis, multiplied this by the proportion of carriers of pathogenic/likely pathogenic variants in the St. Jude Lifetime Cohort (SJLIFE) study, and projected the resulting number onto the U.S. RESULTS: Based on genome sequence data, 11.8% of 2450 SJLIFE participants carry a pathogenic/likely pathogenic variant in one of 156 cancer predisposition genes. Given this information, we estimate that 21 800 adult survivors of childhood cancer in the United States carry a pathogenic/likely pathogenic variant in one of these genes. The highest estimated absolute number of variant carriers are among survivors of central nervous system tumors (n = 4300), particularly astrocytoma (n = 1800) and other gliomas (n = 1700), acute lymphoblastic leukemia (n = 4300), and retinoblastoma (n = 3500). The most frequently mutated genes are RB1 (n = 3000), NF1 (n = 2300), and BRCA2 (n = 800). CONCLUSION: Given the increasing number of childhood cancer survivors in the United States, clinicians should counsel survivors regarding their potential genetic risk, consider referral for genetic counseling and testing, and, as appropriate, implement syndrome-specific cancer surveillance or risk-reducing measures.

Germline Mutations in Predisposition Genes in Pediatric Cancer.

BACKGROUND: The prevalence and spectrum of predisposing mutations among children and adolescents with cancer are largely unknown. Knowledge of such mutations may improve the understanding of tumorigenesis, direct patient care, and enable genetic counseling of patients and families. METHODS: In 1120 patients younger than 20 years of age, we sequenced the whole genomes (in 595 patients), whole exomes (in 456), or both (in 69). We analyzed the DNA sequences of 565 genes, including 60 that have been associated with autosomal dominant cancer-predisposition syndromes, for the presence of germline mutations. The pathogenicity of the mutations was determined by a panel of medical experts with the use of cancer-specific and locus-specific genetic databases, the medical literature, computational predictions, and second hits identified in the tumor genome. The same approach was used to analyze data from 966 persons who did not have known cancer in the 1000 Genomes Project, and a similar approach was used to analyze data from an autism study (from 515 persons with autism and 208 persons without autism). RESULTS: Mutations that were deemed to be pathogenic or probably pathogenic were identified in 95 patients with cancer (8.5%), as compared with 1.1% of the persons in the 1000 Genomes Project and 0.6% of the participants in the autism study. The most commonly mutated genes in the affected patients were TP53 (in 50 patients), APC (in 6), BRCA2 (in 6), NF1 (in 4), PMS2 (in 4), RB1 (in 3), and RUNX1 (in 3). A total of 18 additional patients had protein-truncating mutations in tumor-suppressor genes. Of the 58 patients with a predisposing mutation and available information on family history, 23 (40%) had a family history of cancer. CONCLUSIONS: Germline mutations in cancer-predisposing genes were identified in 8.5% of the children and adolescents with cancer. Family history did not predict the presence of an underlying predisposition syndrome in most patients. (Funded by the American Lebanese Syrian Associated Charities and the National Cancer Institute.).

Germline ETV6 Mutations Confer Susceptibility to Acute Lymphoblastic Leukemia and Thrombocytopenia.

Somatic mutations affecting ETV6 often occur in acute lymphoblastic leukemia (ALL), the most common childhood malignancy. The genetic factors that predispose to ALL remain poorly understood. Here we identify a novel germline ETV6 p. L349P mutation in a kindred affected by thrombocytopenia and ALL. A second ETV6 p. N385fs mutation was identified in an unrelated kindred characterized by thrombocytopenia, ALL and secondary myelodysplasia/acute myeloid leukemia. Leukemic cells from the proband in the second kindred showed deletion of wild type ETV6 with retention of the ETV6 p. N385fs. Enforced expression of the ETV6 mutants revealed normal transcript and protein levels, but impaired nuclear localization. Accordingly, these mutants exhibited significantly reduced ability to regulate the transcription of ETV6 target genes. Our findings highlight a novel role for ETV6 in leukemia predisposition.

Germline genetic variation in ETV6 and risk of childhood acute lymphoblastic leukaemia: a systematic genetic study.

BACKGROUND: Hereditary predisposition is rarely suspected for childhood acute lymphoblastic leukaemia (ALL). Recent reports of germline ETV6 variations associated with substantial familial clustering of haematological malignancies indicated that this gene is a potentially important genetic determinant for ALL susceptibility. Our aims in this study were to comprehensively identify ALL predisposition variants in ETV6 and to determine the extent to which they contributed to the overall risk of childhood ALL. METHODS: Whole-exome sequencing of an index family with several cases of ALL was done to identify causal variants for ALL predisposition. Targeted sequencing of ETV6 was done in children from the Children's Oncology Group and St Jude Children's Research Hospital front-line ALL trials. Patients were included in this study on the basis of their enrolment in these clinical trials and the availability of germline DNA. ETV6 variant genotypes were compared with non-ALL controls to define ALL-related germline risk variants. ETV6 variant function was characterised bioinformatically and correlated with clinical and demographic features in children with ALL. FINDINGS: We identified a novel non-sense ETV6 variant (p.Arg359X) with a high penetrance in an index family. Subsequent targeted sequencing of ETV6 in 4405 childhood ALL cases identified 31 exonic variants (four non-sense, 21 missense, one splice site, and five frameshift variants) that were potentially related to ALL risk in 35 cases (1%). 15 (48%) of 31 ALL-related ETV6 variants clustered in the erythroblast transformation specific domain and were predicted to be highly deleterious. Children with ALL-related ETV6 variants were significantly older at leukaemia diagnosis than those without (10·2 years [IQR 5·3-13·8] vs 4·7 years [3·0-8·7]; p=0·017). The hyperdiploid leukaemia karyotype was highly over-represented in ALL cases harbouring germline ETV6 risk variants compared with the wild-type group (nine [64%] of 14 cases vs 538 [27%] of 2007 cases; p=0·0050). INTERPRETATION: Our findings indicated germline ETV6 variations as the basis of a novel genetic syndrome associated with predisposition to childhood ALL. The development of recommendations for clinical interventions and surveillance for individuals harbouring ALL-related ETV6 variants are needed. FUNDING: US National Institutes of Health and American Lebanese Syrian Associated Charities.

The most informative spacing test effectively discovers biologically relevant outliers or multiple modes in expression.

SUMMARY: Several outlier and subgroup identification statistics (OASIS) have been proposed to discover transcriptomic features with outliers or multiple modes in expression that are indicative of distinct biological processes or subgroups. Here, we borrow ideas from the OASIS methods in the bioinformatics and statistics literature to develop the 'most informative spacing test' (MIST) for unsupervised detection of such transcriptomic features. In an example application involving 14 cases of pediatric acute megakaryoblastic leukemia, MIST more robustly identified features that perfectly discriminate subjects according to gender or the presence of a prognostically relevant fusion-gene than did seven other OASIS methods in the analysis of RNA-seq exon expression, RNA-seq exon junction expression and micorarray exon expression data. MIST was also effective at identifying features related to gender or molecular subtype in an example application involving 157 adult cases of acute myeloid leukemia. AVAILABILITY: MIST will be freely available in the OASIS R package at http://www.stjuderesearch.org/site/depts/biostats CONTACT: stanley.pounds@stjude.org SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Tyrosine kinome sequencing of pediatric acute lymphoblastic leukemia: a report from the Children's Oncology Group TARGET Project.

One recently identified subtype of pediatric B-precursor acute lymphoblastic leukemia (ALL) has been termed BCR-ABL1-like or Ph-like because of similarity of the gene expression profile to BCR-ABL1 positive ALL suggesting the presence of lesions activating tyrosine kinases, frequent alteration of IKZF1, and poor outcome. Prior studies demonstrated that approximately half of these patients had genomic lesions leading to CRLF2 overexpression, with half of such cases harboring somatic mutations in the Janus kinases JAK1 and JAK2. To determine whether mutations in other tyrosine kinases might also occur in ALL, we sequenced the tyrosine kinome and downstream signaling genes in 45 high-risk pediatric ALL cases with either a Ph-like gene expression profile or other alterations suggestive of activated kinase signaling. Aside from JAK mutations and 1 FLT3 mutation, no somatic mutations were found in any other tyrosine kinases, suggesting that alternative mechanisms are responsible for activated kinase signaling in high-risk ALL.

Key pathways are frequently mutated in high-risk childhood acute lymphoblastic leukemia: a report from the Children's Oncology Group.

We sequenced 120 candidate genes in 187 high-risk childhood B-precursor acute lymphoblastic leukemias, the largest pediatric cancer genome sequencing effort reported to date. Integrated analysis of 179 validated somatic sequence mutations with genome-wide copy number alterations and gene expression profiles revealed a high frequency of recurrent somatic alterations in key signaling pathways, including B-cell development/differentiation (68% of cases), the TP53/RB tumor suppressor pathway (54%), Ras signaling (50%), and Janus kinases (11%). Recurrent mutations were also found in ETV6 (6 cases), TBL1XR1 (3), CREBBP (3), MUC4 (2), ASMTL (2), and ADARB2 (2). The frequency of mutations within the 4 major pathways varied markedly across genetic subtypes. Among 23 leukemias expressing a BCR-ABL1-like gene expression profile, 96% had somatic alterations in B-cell development/differentiation, 57% in JAK, and 52% in both pathways, whereas only 9% had Ras pathway mutations. In contrast, 21 cases defined by a distinct gene expression profile coupled with focal ERG deletion rarely had B-cell development/differentiation or JAK kinase alterations but had a high frequency (62%) of Ras signaling pathway mutations. These data extend the range of genes that are recurrently mutated in high-risk childhood B-precursor acute lymphoblastic leukemia and highlight important new therapeutic targets for selected patient subsets.

Defining the condensate landscape of fusion oncoproteins.

Fusion oncoproteins (FOs) arise from chromosomal translocations in ~17% of cancers and are often oncogenic drivers. Although some FOs can promote oncogenesis by undergoing liquid-liquid phase separation (LLPS) to form aberrant biomolecular condensates, the generality of this phenomenon is unknown. We explored this question by testing 166 FOs in HeLa cells and found that 58% formed condensates. The condensate-forming FOs displayed physicochemical features distinct from those of condensate-negative FOs and segregated into distinct feature-based groups that aligned with their sub-cellular localization and biological function. Using Machine Learning, we developed a predictor of FO condensation behavior, and discovered that 67% of ~3000 additional FOs likely form condensates, with 35% of those predicted to function by altering gene expression. 47% of the predicted condensate-negative FOs were associated with cell signaling functions, suggesting a functional dichotomy between condensate-positive and -negative FOs. Our Datasets and reagents are rich resources to interrogate FO condensation in the future.

Bambino: a variant detector and alignment viewer for next-generation sequencing data in the SAM/BAM format.

SUMMARY: Bambino is a variant detector and graphical alignment viewer for next-generation sequencing data in the SAM/BAM format, which is capable of pooling data from multiple source files. The variant detector takes advantage of SAM-specific annotations, and produces detailed output suitable for genotyping and identification of somatic mutations. The assembly viewer can display reads in the context of either a user-provided or automatically generated reference sequence, retrieve genome annotation features from a UCSC genome annotation database, display histograms of non-reference allele frequencies, and predict protein-coding changes caused by SNPs. AVAILABILITY: Bambino is written in platform-independent Java and available from https://cgwb.nci.nih.gov/goldenPath/bamview/documentation/index.html, along with documentation and example data. Bambino may be launched online via Java Web Start or downloaded and run locally.

Exploration of Coding and Non-coding Variants in Cancer Using GenomePaint.

GenomePaint (https://genomepaint.stjude.cloud/) is an interactive visualization platform for whole-genome, whole-exome, transcriptome, and epigenomic data of tumor samples. Its design captures the inter-relatedness between DNA variations and RNA expression, supporting in-depth exploration of both individual cancer genomes and full cohorts. Regulatory non-coding variants can be inspected and analyzed along with coding variants, and their functional impact further explored by examining 3D genome data from cancer cell lines. Further, GenomePaint correlates mutation and expression patterns with patient outcomes, and supports custom data upload. We used GenomePaint to unveil aberrant splicing that disrupts the RING domain of CREBBP, discover cis activation of the MYC oncogene by duplication of the NOTCH1-MYC enhancer in B-lineage acute lymphoblastic leukemia, and explore the inter- and intra-tumor heterogeneity at EGFR in adult glioblastomas. These examples demonstrate that deep multi-omics exploration of individual cancer genomes enabled by GenomePaint can lead to biological insights for follow-up validation.