Multiplatform analysis of 12 cancer types reveals molecular classification within and across tissues of origin.

Recent genomic analyses of pathologically defined tumor types identify "within-a-tissue" disease subtypes. However, the extent to which genomic signatures are shared across tissues is still unclear. We performed an integrative analysis using five genome-wide platforms and one proteomic platform on 3,527 specimens from 12 cancer types, revealing a unified classification into 11 major subtypes. Five subtypes were nearly identical to their tissue-of-origin counterparts, but several distinct cancer types were found to converge into common subtypes. Lung squamous, head and neck, and a subset of bladder cancers coalesced into one subtype typified by TP53 alterations, TP63 amplifications, and high expression of immune and proliferation pathway genes. Of note, bladder cancers split into three pan-cancer subtypes. The multiplatform classification, while correlated with tissue-of-origin, provides independent information for predicting clinical outcomes. All data sets are available for data-mining from a unified resource to support further biological discoveries and insights into novel therapeutic strategies.

An open protocol for modeling T Cell Clonotype repertoires using TCRß CDR3 sequences.

T cell receptor repertoires can be profiled using next generation sequencing (NGS) to measure and monitor adaptive dynamical changes in response to disease and other perturbations. Genomic DNA-based bulk sequencing is cost-effective but necessitates multiplex target amplification using multiple primer pairs with highly variable amplification efficiencies. Here, we utilize an equimolar primer mixture and propose a single statistical normalization step that efficiently corrects for amplification bias post sequencing. Using samples analyzed by both our open protocol and a commercial solution, we show high concordance between bulk clonality metrics. This approach is an inexpensive and open-source alternative to commercial solutions.

svpluscnv: analysis and visualization of complex structural variation data.

MOTIVATION: Despite widespread prevalence of somatic structural variations (SVs) across most tumor types, understanding of their molecular implications often remains poor. SVs are extremely heterogeneous in size and complexity, hindering the interpretation of their pathogenic role. Tools integrating large SV datasets across platforms are required to fully characterize the cancer's somatic landscape. RESULTS: svpluscnv R package is a swiss army knife for the integration and interpretation of orthogonal datasets including copy number variant segmentation profiles and sequencing-based structural variant calls. The package implements analysis and visualization tools to evaluate chromosomal instability and ploidy, identify genes harboring recurrent SVs and detects complex rearrangements such as chromothripsis and chromoplexia. Further, it allows systematic identification of hot-spot shattered genomic regions, showing reproducibility across alternative detection methods and datasets. AVAILABILITY AND IMPLEMENTATION: https://github.com/ccbiolab/svpluscnv. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Pan-cancer transcriptional signatures predictive of oncogenic mutations reveal that Fbw7 regulates cancer cell oxidative metabolism.

The Fbw7 (F-box/WD repeat-containing protein 7) ubiquitin ligase targets multiple oncoproteins for degradation and is commonly mutated in cancers. Like other pleiotropic tumor suppressors, Fbw7's complex biology has impeded our understanding of how Fbw7 mutations promote tumorigenesis and hindered the development of targeted therapies. To address these needs, we employed a transfer learning approach to derive gene-expression signatures from The Cancer Gene Atlas datasets that predict Fbw7 mutational status across tumor types and identified the pathways enriched within these signatures. Genes involved in mitochondrial function were highly enriched in pan-cancer signatures that predict Fbw7 mutations. Studies in isogenic colorectal cancer cell lines that differed in Fbw7 mutational status confirmed that Fbw7 mutations increase mitochondrial gene expression. Surprisingly, Fbw7 mutations shifted cellular metabolism toward oxidative phosphorylation and caused context-specific metabolic vulnerabilities. Our approach revealed unexpected metabolic reprogramming and possible therapeutic targets in Fbw7-mutant cancers and provides a framework to study other complex, oncogenic mutations.

Valection: design optimization for validation and verification studies.

BACKGROUND: Platform-specific error profiles necessitate confirmatory studies where predictions made on data generated using one technology are additionally verified by processing the same samples on an orthogonal technology. However, verifying all predictions can be costly and redundant, and testing a subset of findings is often used to estimate the true error profile. RESULTS: To determine how to create subsets of predictions for validation that maximize accuracy of global error profile inference, we developed Valection, a software program that implements multiple strategies for the selection of verification candidates. We evaluated these selection strategies on one simulated and two experimental datasets. CONCLUSIONS: Valection is implemented in multiple programming languages, available at: http://labs.oicr.on.ca/boutros-lab/software/valection.

Germline contamination and leakage in whole genome somatic single nucleotide variant detection.

BACKGROUND: The clinical sequencing of cancer genomes to personalize therapy is becoming routine across the world. However, concerns over patient re-identification from these data lead to questions about how tightly access should be controlled. It is not thought to be possible to re-identify patients from somatic variant data. However, somatic variant detection pipelines can mistakenly identify germline variants as somatic ones, a process called "germline leakage". The rate of germline leakage across different somatic variant detection pipelines is not well-understood, and it is uncertain whether or not somatic variant calls should be considered re-identifiable. To fill this gap, we quantified germline leakage across 259 sets of whole-genome somatic single nucleotide variant (SNVs) predictions made by 21 teams as part of the ICGC-TCGA DREAM Somatic Mutation Calling Challenge. RESULTS: The median somatic SNV prediction set contained 4325 somatic SNVs and leaked one germline polymorphism. The level of germline leakage was inversely correlated with somatic SNV prediction accuracy and positively correlated with the amount of infiltrating normal cells. The specific germline variants leaked differed by tumour and algorithm. To aid in quantitation and correction of leakage, we created a tool, called GermlineFilter, for use in public-facing somatic SNV databases. CONCLUSIONS: The potential for patient re-identification from leaked germline variants in somatic SNV predictions has led to divergent open data access policies, based on different assessments of the risks. Indeed, a single, well-publicized re-identification event could reshape public perceptions of the values of genomic data sharing. We find that modern somatic SNV prediction pipelines have low germline-leakage rates, which can be further reduced, especially for cloud-sharing, using pre-filtering software.

Combining tumor genome simulation with crowdsourcing to benchmark somatic single-nucleotide-variant detection.

The detection of somatic mutations from cancer genome sequences is key to understanding the genetic basis of disease progression, patient survival and response to therapy. Benchmarking is needed for tool assessment and improvement but is complicated by a lack of gold standards, by extensive resource requirements and by difficulties in sharing personal genomic information. To resolve these issues, we launched the ICGC-TCGA DREAM Somatic Mutation Calling Challenge, a crowdsourced benchmark of somatic mutation detection algorithms. Here we report the BAMSurgeon tool for simulating cancer genomes and the results of 248 analyses of three in silico tumors created with it. Different algorithms exhibit characteristic error profiles, and, intriguingly, false positives show a trinucleotide profile very similar to one found in human tumors. Although the three simulated tumors differ in sequence contamination (deviation from normal cell sequence) and in subclonality, an ensemble of pipelines outperforms the best individual pipeline in all cases. BAMSurgeon is available at https://github.com/adamewing/bamsurgeon/.

Modeling gene-wise dependencies improves the identification of drug response biomarkers in cancer studies.

Motivation: In recent years, vast advances in biomedical technologies and comprehensive sequencing have revealed the genomic landscape of common forms of human cancer in unprecedented detail. The broad heterogeneity of the disease calls for rapid development of personalized therapies. Translating the readily available genomic data into useful knowledge that can be applied in the clinic remains a challenge. Computational methods are needed to aid these efforts by robustly analyzing genome-scale data from distinct experimental platforms for prioritization of targets and treatments. Results: We propose a novel, biologically motivated, Bayesian multitask approach, which explicitly models gene-centric dependencies across multiple and distinct genomic platforms. We introduce a gene-wise prior and present a fully Bayesian formulation of a group factor analysis model. In supervised prediction applications, our multitask approach leverages similarities in response profiles of groups of drugs that are more likely to be related to true biological signal, which leads to more robust performance and improved generalization ability. We evaluate the performance of our method on molecularly characterized collections of cell lines profiled against two compound panels, namely the Cancer Cell Line Encyclopedia and the Cancer Therapeutics Response Portal. We demonstrate that accounting for the gene-centric dependencies enables leveraging information from multi-omic input data and improves prediction and feature selection performance. We further demonstrate the applicability of our method in an unsupervised dimensionality reduction application by inferring genes essential to tumorigenesis in the pancreatic ductal adenocarcinoma and lung adenocarcinoma patient cohorts from The Cancer Genome Atlas. Availability and Implementation: : The code for this work is available at https://github.com/olganikolova/gbgfa. Contact: : nikolova@ohsu.edu or margolin@ohsu.edu. Supplementary information: Supplementary data are available at Bioinformatics online.

Population-level distribution and putative immunogenicity of cancer neoepitopes.

BACKGROUND: Tumor neoantigens are drivers of cancer immunotherapy response; however, current prediction tools produce many candidates requiring further prioritization. Additional filtration criteria and population-level understanding may assist with prioritization. Herein, we show neoepitope immunogenicity is related to measures of peptide novelty and report population-level behavior of these and other metrics. METHODS: We propose four peptide novelty metrics to refine predicted neoantigenicity: tumor vs. paired normal peptide binding affinity difference, tumor vs. paired normal peptide sequence similarity, tumor vs. closest human peptide sequence similarity, and tumor vs. closest microbial peptide sequence similarity. We apply these metrics to neoepitopes predicted from somatic missense mutations in The Cancer Genome Atlas (TCGA) and a cohort of melanoma patients, and to a group of peptides with neoepitope-specific immune response data using an extension of pVAC-Seq (Hundal et al., pVAC-Seq: a genome-guided in silico approach to identifying tumor neoantigens. Genome Med 8:11, 2016). RESULTS: We show neoepitope burden varies across TCGA diseases and HLA alleles, with surprisingly low repetition of neoepitope sequences across patients or neoepitope preferences among sets of HLA alleles. Only 20.3% of predicted neoepitopes across TCGA patients displayed novel binding change based on our binding affinity difference criteria. Similarity of amino acid sequence was typically high between paired tumor-normal epitopes, but in 24.6% of cases, neoepitopes were more similar to other human peptides, or bacterial (56.8% of cases) or viral peptides (15.5% of cases), than their paired normal counterparts. Applied to peptides with neoepitope-specific immune response, a linear model incorporating neoepitope binding affinity, protein sequence similarity between neoepitopes and their closest viral peptides, and paired binding affinity difference was able to predict immunogenicity (AUROC = 0.66). CONCLUSIONS: Our proposed prioritization criteria emphasize neoepitope novelty and refine patient neoepitope predictions for focus on biologically meaningful candidate neoantigens. We have demonstrated that neoepitopes should be considered not only with respect to their paired normal epitope, but to the entire human proteome, and bacterial and viral peptides, with potential implications for neoepitope immunogenicity and personalized vaccines for cancer treatment. We conclude that putative neoantigens are highly variable across individuals as a function of cancer genetics and personalized HLA repertoire, while the overall behavior of filtration criteria reflects predictable patterns.

The Cancer Cell Line Encyclopedia enables predictive modelling of anticancer drug sensitivity.

The systematic translation of cancer genomic data into knowledge of tumour biology and therapeutic possibilities remains challenging. Such efforts should be greatly aided by robust preclinical model systems that reflect the genomic diversity of human cancers and for which detailed genetic and pharmacological annotation is available. Here we describe the Cancer Cell Line Encyclopedia (CCLE): a compilation of gene expression, chromosomal copy number and massively parallel sequencing data from 947 human cancer cell lines. When coupled with pharmacological profiles for 24 anticancer drugs across 479 of the cell lines, this collection allowed identification of genetic, lineage, and gene-expression-based predictors of drug sensitivity. In addition to known predictors, we found that plasma cell lineage correlated with sensitivity to IGF1 receptor inhibitors; AHR expression was associated with MEK inhibitor efficacy in NRAS-mutant lines; and SLFN11 expression predicted sensitivity to topoisomerase inhibitors. Together, our results indicate that large, annotated cell-line collections may help to enable preclinical stratification schemata for anticancer agents. The generation of genetic predictions of drug response in the preclinical setting and their incorporation into cancer clinical trial design could speed the emergence of 'personalized' therapeutic regimens.

ATARiS: computational quantification of gene suppression phenotypes from multisample RNAi screens.

Genome-scale RNAi libraries enable the systematic interrogation of gene function. However, the interpretation of RNAi screens is complicated by the observation that RNAi reagents designed to suppress the mRNA transcripts of the same gene often produce a spectrum of phenotypic outcomes due to differential on-target gene suppression or perturbation of off-target transcripts. Here we present a computational method, Analytic Technique for Assessment of RNAi by Similarity (ATARiS), that takes advantage of patterns in RNAi data across multiple samples in order to enrich for RNAi reagents whose phenotypic effects relate to suppression of their intended targets. By summarizing only such reagent effects for each gene, ATARiS produces quantitative, gene-level phenotype values, which provide an intuitive measure of the effect of gene suppression in each sample. This method is robust for data sets that contain as few as 10 samples and can be used to analyze screens of any number of targeted genes. We used this analytic approach to interrogate RNAi data derived from screening more than 100 human cancer cell lines and identified HNF1B as a transforming oncogene required for the survival of cancer cells that harbor HNF1B amplifications. ATARiS is publicly available at http://broadinstitute.org/ataris.

Drug susceptibility prediction against a panel of drugs using kernelized Bayesian multitask learning.

MOTIVATION: Human immunodeficiency virus (HIV) and cancer require personalized therapies owing to their inherent heterogeneous nature. For both diseases, large-scale pharmacogenomic screens of molecularly characterized samples have been generated with the hope of identifying genetic predictors of drug susceptibility. Thus, computational algorithms capable of inferring robust predictors of drug responses from genomic information are of great practical importance. Most of the existing computational studies that consider drug susceptibility prediction against a panel of drugs formulate a separate learning problem for each drug, which cannot make use of commonalities between subsets of drugs. RESULTS: In this study, we propose to solve the problem of drug susceptibility prediction against a panel of drugs in a multitask learning framework by formulating a novel Bayesian algorithm that combines kernel-based non-linear dimensionality reduction and binary classification (or regression). The main novelty of our method is the joint Bayesian formulation of projecting data points into a shared subspace and learning predictive models for all drugs in this subspace, which helps us to eliminate off-target effects and drug-specific experimental noise. Another novelty of our method is the ability of handling missing phenotype values owing to experimental conditions and quality control reasons. We demonstrate the performance of our algorithm via cross-validation experiments on two benchmark drug susceptibility datasets of HIV and cancer. Our method obtains statistically significantly better predictive performance on most of the drugs compared with baseline single-task algorithms that learn drug-specific models. These results show that predicting drug susceptibility against a panel of drugs simultaneously within a multitask learning framework improves overall predictive performance over single-task learning approaches. AVAILABILITY AND IMPLEMENTATION: Our Matlab implementations for binary classification and regression are available at https://github.com/mehmetgonen/kbmtl. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Improving breast cancer survival analysis through competition-based multidimensional modeling.

Breast cancer is the most common malignancy in women and is responsible for hundreds of thousands of deaths annually. As with most cancers, it is a heterogeneous disease and different breast cancer subtypes are treated differently. Understanding the difference in prognosis for breast cancer based on its molecular and phenotypic features is one avenue for improving treatment by matching the proper treatment with molecular subtypes of the disease. In this work, we employed a competition-based approach to modeling breast cancer prognosis using large datasets containing genomic and clinical information and an online real-time leaderboard program used to speed feedback to the modeling team and to encourage each modeler to work towards achieving a higher ranked submission. We find that machine learning methods combined with molecular features selected based on expert prior knowledge can improve survival predictions compared to current best-in-class methodologies and that ensemble models trained across multiple user submissions systematically outperform individual models within the ensemble. We also find that model scores are highly consistent across multiple independent evaluations. This study serves as the pilot phase of a much larger competition open to the whole research community, with the goal of understanding general strategies for model optimization using clinical and molecular profiling data and providing an objective, transparent system for assessing prognostic models.

Reverse engineering of regulatory networks in human B cells.

Cellular phenotypes are determined by the differential activity of networks linking coregulated genes. Available methods for the reverse engineering of such networks from genome-wide expression profiles have been successful only in the analysis of lower eukaryotes with simple genomes. Using a new method called ARACNe (algorithm for the reconstruction of accurate cellular networks), we report the reconstruction of regulatory networks from expression profiles of human B cells. The results are suggestive a hierarchical, scale-free network, where a few highly interconnected genes (hubs) account for most of the interactions. Validation of the network against available data led to the identification of MYC as a major hub, which controls a network comprising known target genes as well as new ones, which were biochemically validated. The newly identified MYC targets include some major hubs. This approach can be generally useful for the analysis of normal and pathologic networks in mammalian cells.

An open protocol for modeling T Cell Clonotype repertoires using TCRß CDR3 sequences.

T cell receptor repertoires can be profiled using next generation sequencing (NGS) to measure and monitor adaptive dynamical changes in response to disease and other perturbations. Genomic DNA-based bulk sequencing is cost-effective but necessitates multiplex target amplification using multiple primer pairs with highly variable amplification efficiencies. Here, we utilize an equimolar primer mixture and propose a single statistical normalization step that efficiently corrects for amplification bias post sequencing. Using samples analyzed by both our open protocol and a commercial solution, we show high concordance between bulk clonality metrics. This approach is an inexpensive and open-source alternative to commercial solutions.

Integrated biochemical and computational approach identifies BCL6 direct target genes controlling multiple pathways in normal germinal center B cells.

BCL6 is a transcriptional repressor required for mature B-cell germinal center (GC) formation and implicated in lymphomagenesis. BCL6's physiologic function is only partially known because the complete set of its targets in GC B cells has not been identified. To address this issue, we used an integrated biochemical-computational-functional approach to identify BCL6 direct targets in normal GC B cells. This approach includes (1) identification of BCL6-bound promoters by genome-wide chromatin immunoprecipitation, (2) inference of transcriptional relationships by the use of a regulatory network reverse engineering approach (ARACNe), and (3) validation of physiologic relevance of the candidate targets down-regulated in GC B cells. Our approach demonstrated that a large set of promoters (> 4000) is physically bound by BCL6 but that only a fraction of them is repressed in GC B cells. This set of 1207 targets identifies several cellular functions directly controlled by BCL6 during GC development, including activation, survival, DNA-damage response, cell cycle arrest, cytokine signaling, Toll-like receptor signaling, and differentiation. These results define a broad role of BCL6 in preventing centroblasts from responding to signals leading to exit from the GC before they complete the phase of proliferative expansion and of antibody affinity maturation.

ChIP-on-chip significance analysis reveals large-scale binding and regulation by human transcription factor oncogenes.

ChIP-on-chip has emerged as a powerful tool to dissect the complex network of regulatory interactions between transcription factors and their targets. However, most ChIP-on-chip analysis methods use conservative approaches aimed at minimizing false-positive transcription factor targets. We present a model with improved sensitivity in detecting binding events from ChIP-on-chip data. Its application to human T cells, followed by extensive biochemical validation, reveals that 3 oncogenic transcription factors, NOTCH1, MYC, and HES1, bind to several thousand target gene promoters, up to an order of magnitude increase over conventional analysis methods. Gene expression profiling upon NOTCH1 inhibition shows broad-scale functional regulation across the entire range of predicted target genes, establishing a closer link between occupancy and regulation. Finally, the increased sensitivity reveals a combinatorial regulatory program in which MYC cobinds to virtually all NOTCH1-bound promoters. Overall, these results suggest an unappreciated complexity of transcriptional regulatory networks and highlight the fundamental importance of genome-scale analysis to represent transcriptional programs.

Identifying the proteins to which small-molecule probes and drugs bind in cells.

Most small-molecule probes and drugs alter cell circuitry by interacting with 1 or more proteins. A complete understanding of the interacting proteins and their associated protein complexes, whether the compounds are discovered by cell-based phenotypic or target-based screens, is extremely rare. Such a capability is expected to be highly illuminating--providing strong clues to the mechanisms used by small-molecules to achieve their recognized actions and suggesting potential unrecognized actions. We describe a powerful method combining quantitative proteomics (SILAC) with affinity enrichment to provide unbiased, robust and comprehensive identification of the proteins that bind to small-molecule probes and drugs. The method is scalable and general, requiring little optimization across different compound classes, and has already had a transformative effect on our studies of small-molecule probes. Here, we describe in full detail the application of the method to identify targets of kinase inhibitors and immunophilin binders.

A harmonized meta-knowledgebase of clinical interpretations of somatic genomic variants in cancer.

Precision oncology relies on accurate discovery and interpretation of genomic variants, enabling individualized diagnosis, prognosis and therapy selection. We found that six prominent somatic cancer variant knowledgebases were highly disparate in content, structure and supporting primary literature, impeding consensus when evaluating variants and their relevance in a clinical setting. We developed a framework for harmonizing variant interpretations to produce a meta-knowledgebase of 12,856 aggregate interpretations. We demonstrated large gains in overlap between resources across variants, diseases and drugs as a result of this harmonization. We subsequently demonstrated improved matching between a patient cohort and harmonized interpretations of potential clinical significance, observing an increase from an average of 33% per individual knowledgebase to 57% in aggregate. Our analyses illuminate the need for open, interoperable sharing of variant interpretation data. We also provide a freely available web interface (search.cancervariants.org) for exploring the harmonized interpretations from these six knowledgebases.