Tumor Sequencing Projects Research Articles

Abstract B032: Evaluating T cell-inflammation in pancreatic neuroendocrine tumors

Abstract Background: Pancreatic neuroendocrine tumors (pNETs) are rare tumors arising from neuroendocrine cells in the pancreas. Their clinical behaviour and prognosis are variable and durable responses are often difficult to achieve with conventional therapies. Immunotherapy approaches may be efficacious in a subgroup of pNETs. In a recent study that did not include pNETs, we showed that a four-chemokine signature (c-Score: CCL4, CCL5, CXCL9, CXCL10) predicts T cell-inflammation and response to immune checkpoint inhibitors across cancers. We have hypothesized that the c-Score will also identify pNET subgroups with T cell-inflammation. Methods: The c-Score and hallmarks of the cancer-immunity cycle are being evaluated in pNETs as a cohort study (n=100 cases) of the Marathon of Hope Cancer Centres Network whole genome and transcriptome tumor sequencing (WGTS) project. The cohort study will compare pNET profiles with those of other pancreatic and peri-ampullary tumors. Results: As of June 1, 2024, primary tumor data from 23 cases were available for analysis, including 16 pNETs, 2 pancreatic ductal adenocarcinomas (PDAC), 2 pancreatic adenosquamous carcinoma (PASC), 2 distal cholangiocarcinomas (dCCA), and 1 ampullary carcinoma (AC). Dimension reduction by principle component analysis (PCA) revealed clustering based on histological subtype. Interestingly, PCA using immune genes also demonstrate subgrouping that overlapped with the histological subtypes, suggesting that immunity of pancreatic/peri-ampullary neoplasms may be directed by histological types rather than the pancreas gland itself. These cancers showed a broad distribution of c-Score expression, with pNETs having the lowest c-Score. Tumors with high expression of the c-Score, including 2 pNETs, had increased expression of cancer-immunity cycle gene sets. DUX4, FAM186A, and IGFN1 were amongst the most common genes mutated in all tumors. Conclusions: Early data suggest that pNETs generally harbor low c-Score values and hallmarks of cancer-immunity, with the existence of a subgroup of pNETs displaying T cell-inflammation. Citation Format: Joan Miguel Romero, James Vafiadis, Tess Aalto, Ruomeng Fang, Yifan Wang, George Zogopoulos. Evaluating T cell-inflammation in pancreatic neuroendocrine tumors [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pancreatic Cancer Research; 2024 Sep 15-18; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl_2):Abstract nr B032.

The Landscape of the Heritable Cancer Genome

Genome-wide association studies (GWAS) have found hundreds of single-nucleotide polymorphisms (SNP) associated with increased risk of cancer. However, the amount of heritable risk explained by SNPs is limited, leaving most of the cancer heritability unexplained. Tumor sequencing projects have shown that causal mutations are enriched in genic regions. We hypothesized that SNPs located in protein coding genes and nearby regulatory regions could explain a significant proportion of the heritable risk of cancer. To perform gene-level heritability analysis, we developed a new method, called Bayesian Gene Heritability Analysis (BAGHERA), to estimate the heritability explained by all genotyped SNPs and by those located in genic regions using GWAS summary statistics. BAGHERA was specifically designed for low heritability traits such as cancer and provides robust heritability estimates under different genetic architectures. BAGHERA-based analysis of 38 cancers reported in the UK Biobank showed that SNPs explain at least 10% of the heritable risk for 14 of them, including late onset malignancies. We then identified 1,146 genes, called cancer heritability genes (CHG), explaining a significant proportion of cancer heritability. CHGs were involved in hallmark processes controlling the transformation from normal to cancerous cells. Importantly, 60 of them also harbored somatic driver mutations, and 27 are tumor suppressors. Our results suggest that germline and somatic mutation information could be exploited to identify subgroups of individuals at higher risk of cancer in the broader population and could prove useful to establish strategies for early detection and cancer surveillance. SIGNIFICANCE: This study describes a new statistical method to identify genes associated with cancer heritability in the broader population, creating a map of the heritable cancer genome with gene-level resolution.See related commentary by Bader, p. 2586.

MuSE: accounting for tumor heterogeneity using a sample-specific error model improves sensitivity and specificity in mutation calling from sequencing data.

Subclonal mutations reveal important features of the genetic architecture of tumors. However, accurate detection of mutations in genetically heterogeneous tumor cell populations using next-generation sequencing remains challenging. We develop MuSE (http://bioinformatics.mdanderson.org/main/MuSE), Mutation calling using a Markov Substitution model for Evolution, a novel approach for modeling the evolution of the allelic composition of the tumor and normal tissue at each reference base. MuSE adopts a sample-specific error model that reflects the underlying tumor heterogeneity to greatly improve the overall accuracy. We demonstrate the accuracy of MuSE in calling subclonal mutations in the context of large-scale tumor sequencing projects using whole exome and whole genome sequencing.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-016-1029-6) contains supplementary material, which is available to authorized users.

Abstract A2-42: Identifying clinically important somatic mutations through a knowledge-based approach

Abstract Large-scale tumor sequencing projects, like The Cancer Genome Atlas (TCGA), have implicated thousands of somatic mutations in cancer. These initiatives have incentivized many improvements in somatic variant detection. However, we have observed that important pathogenic variants are often missed due to stringent filtering, tumor heterogeneity, tumor contamination of normal, low tumor purity, alignment challenges, and other issues. These idiosyncrasies can impede variant detection algorithms from reliably calling even the most clinically relevant variants. To rescue this missed variation we devised a knowledge based variant identification strategy. We mined the literature and other variant databases for pathogenic variation and assembled them into an integrated Database of Curated Mutations(DoCM - www.docm.info). The DoCM contains 488 variants across 63 genes implicated in 34 cancer types. We developed an algorithm to identify any pathogenic variant signal, for all variants in the DoCM, in aligned sequence data. As a proof of principle, we applied this approach to four cancer types sequenced by TCGA: acute myeloid leukemia (AML), breast cancer, ovarian carcinoma, and uterine corpus endometrial carcinoma. Obvious sequencing and alignment errors, like variants in homopolymer runs, were excluded from subsequent analysis by manual review. Across these four TCGA projects, which includes 1,840 individuals, 1,757 clinically relevant variants were identified, 1,223 of which had not been previously reported in TCGA studies. To validate this approach, custom capture probes were designed for all of the DoCM variants, new libraries constructed and deep sequencing performed on 96 tumor and matched normal samples from the AML and breast cancer TCGA projects. Following this strategy, we were able to confirm the rescue of clinically relevant somatic mutations that were missed in the original TCGA analysis. We propose a knowledge-driven variant detection approach be considered as standard practice to avoid false-negative calls of events likely to be clinically relevant. Citation Format: Benjamin J. Ainscough, Malachi Griffith, Jason Kunisaki, Adam Coffman, Joshua F. McMichael, James M. Eldred, Jason R. Walker, Robert S. Fulton, Richard K. Wilson, Obi L. Griffith, Elaine R. Mardis. Identifying clinically important somatic mutations through a knowledge-based approach. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr A2-42.

Abstract PR01: Identifying clinically important somatic mutations through a knowledge-based approach

Abstract Large-scale tumor sequencing projects, like The Cancer Genome Atlas (TCGA), have implicated thousands of somatic mutations in cancer. These initiatives have incentivized many improvements in somatic variant detection. However, we have observed that important pathogenic variants are often missed due to stringent filtering, tumor heterogeneity, tumor contamination of normal, low tumor purity, alignment challenges, and other issues. These idiosyncrasies can impede variant detection algorithms from reliably calling even the most clinically relevant variants. To rescue this missed variation we devised a knowledge based variant identification strategy. We mined the literature and other variant databases for pathogenic variation and assembled them into an integrated Database of Curated Mutations (DoCM - www.docm.info). The DoCM contains 488 variants across 63 genes implicated in 34 cancer types. We developed an algorithm to identify any pathogenic variant signal, for all variants in the DoCM, in aligned sequence data. As a proof of principle, we applied this approach to four cancer types sequenced by TCGA: acute myeloid leukemia (AML), breast cancer, ovarian carcinoma, and uterine corpus endometrial carcinoma. Obvious sequencing and alignment errors, like variants in homopolymer runs, were excluded from subsequent analysis by manual review. Across these four TCGA projects, which includes 1,840 individuals, 1,757 clinically relevant variants were identified, 1,223 of which had not been previously reported in TCGA studies. To validate this approach, custom capture probes were designed for all of the DoCM variants, new libraries constructed and deep sequencing performed on 96 tumor and matched normal samples from the AML and breast cancer TCGA projects. Following this strategy, we were able to confirm the rescue of clinically relevant somatic mutations that were missed in the original TCGA analysis. We propose a knowledge-driven variant detection approach be considered as standard practice to avoid false-negative calls of events likely to be clinically relevant Citation Format: Benjamin J. Ainscough, Malachi Griffith, Jason Kunisaki, Adam Coffman, Joshua F. McMichael, James M. Eldred, Jason R. Walker, Robert S. Fulton, Richard K. Wilson, Obi L. Griffith, Elaine R. Mardis. Identifying clinically important somatic mutations through a knowledge-based approach. [abstract]. In: Proceedings of the AACR Special Conference on Translation of the Cancer Genome; Feb 7-9, 2015; San Francisco, CA. Philadelphia (PA): AACR; Cancer Res 2015;75(22 Suppl 1):Abstract nr PR01.

Adjusting for Background Mutation Frequency Biases Improves the Identification of Cancer Driver Genes

A common goal of tumor sequencing projects is finding genes whose mutations are selected for during tumor development. This is accomplished by choosing genes that have more non-synonymous mutations than expected from an estimated background mutation frequency. While this background frequency is unknown, it can be estimated using both the observed synonymous mutation frequency and the non-synonymous to synonymous mutation ratio. The synonymous mutation frequency can be determined across all genes or in a gene-specific manner. This choice introduces an interesting trade-off. A gene-specific frequency adjusts for an underlying mutation bias, but is difficult to estimate given missing synonymous mutation counts. Using a genome-wide synonymous frequency is more robust, but is less suited for adjusting biases. Studying four evaluation criteria for identifying genes with high non-synonymous mutation burden (reflecting preferential selection of expressed genes, genes with mutations in conserved bases, genes with many protein interactions, and genes that show loss of heterozygosity), we find that the gene-specific synonymous frequency is superior in the gene expression and protein interaction tests. In conclusion, the use of the gene-specific synonymous mutation frequency is well suited for assessing a gene's non-synonymous mutation burden.

Abstract 5141: GenomeSpace: an environment for frictionless bioinformatics.

Abstract Over the past several years, cancer genome characterization initiatives such as The Cancer Genome Atlas, International Cancer Genome Consortium, and the Tumor Sequencing Project have produced an explosion of genomic data. The pace of data production has increased with the adoption of next-generation sequencing technologies and large-scale data production efforts to discover the breadth of genomic variation in humans. Comprehensive analysis of these datasets requires the coordinated use of Web-based data repositories and applications, desktop analysis tools and visualizers, and single-purpose algorithms. However, the effort required to transfer data between tools, convert between data formats, and manage results often prevents researchers from utilizing the wealth of methods available to them. Many integrative genomics and translational “bench to bedside” discoveries are possible with combinations of existing tools, but the necessary transitions between them puts them out of the reach of most researchers. Cloud technologies have produced a new wave of applications that transfer next-generation sequence data directly from sequencers to a compute cloud for storage and analysis, but these systems are at an early stage of maturity and many are tied to commercial sequencing vendors. GenomeSpace, http://www.genomespace.org, is an environment that brings together diverse computational tools, enabling scientists without programming skills to easily combine their capabilities. It aims to offer a common space to create, manipulate and share an ever-growing range of genomic analysis tools. GenomeSpace features support for cloud-based data storage and analysis, multi-tool analytic workflows, automatic conversion of data formats, and ease of connecting new tools to the environment. A set of six “GenomeSpace-enabled” seed tools developed by collaborating organizations provides a comprehensive platform for the analysis of cancer data: Cytoscape (UCSD), Galaxy (Penn State University), GenePattern (Broad Institute), Genomica (Weitzmann Institute), Integrative Genomics Viewer (Broad Institute), and the UCSC Genome Browser (UCSC). The extensible format of the system has empowered a wider range of cancer analyses through the addition of ArrayExpress (European Bioinformatics Institute), InSilico DB (University of Brussels), geWorkbench (Columbia University), and Cistrome (Dana-Farber Cancer Institute). We show how researchers can use GenomeSpace to effortlessly combine the capabilities of all of these tools in several cancer research scenarios. Citation Format: Michael Reich, John Liefeld, Helga Thorvaldsdottir, Marco Ocana, Thorin Tabor, DK Jang, Jill P. Mesirov. GenomeSpace: an environment for frictionless bioinformatics. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 5141. doi:10.1158/1538-7445.AM2013-5141

Domain landscapes of somatic mutations in cancer

BackgroundLarge-scale tumor sequencing projects are now underway to identify genetic mutations that drive tumor initiation and development. Most studies take a gene-based approach to identifying driver mutations, highlighting genes mutated in a large percentage of tumor samples as those likely to contain driver mutations. However, this gene-based approach usually does not consider the position of the mutation within the gene or the functional context the position of the mutation provides. Here we introduce a novel method for mapping mutations to distinct protein domains, not just individual genes, in which they occur, thus providing the functional context for how the mutation contributes to disease. Furthermore, aggregating mutations from all genes containing a specific protein domain enables the identification of mutations that are rare at the gene level, but that occur frequently within the specified domain. These highly mutated domains potentially reveal disruptions of protein function necessary for cancer development.ResultsWe mapped somatic mutations from the protein coding regions of 100 colon adenocarcinoma tumor samples to the genes and protein domains in which they occurred, and constructed topographical maps to depict the “mutational landscapes” of gene and domain mutation frequencies. We found significant mutation frequency in a number of genes previously known to be somatically mutated in colon cancer patients including APC, TP53 and KRAS. In addition, we found significant mutation frequency within specific domains located in these genes, as well as within other domains contained in genes having low mutation frequencies. These domain “peaks” were enriched with functions important to cancer development including kinase activity, DNA binding and repair, and signal transduction.ConclusionsUsing our method to create the domain landscapes of mutations in colon cancer, we were able to identify somatic mutations with high potential to drive cancer development. Interestingly, the majority of the genes involved have a low mutation frequency. Therefore, themethod shows good potential for identifying rare driver mutations in current, large-scale tumor sequencing projects. In addition, mapping mutations to specific domains provides the necessary functional context for understanding how the mutations contribute to the disease, and may reveal novel or more refined gene and domain target regions for drug development.

Abstract 3966: GenomeSpace: An environment for frictionless bioinformatics

Abstract Over the past several years, cancer genome characterization initiatives such as The Cancer Genome Atlas and the Tumor Sequencing Project have produced an explosion of genomic data. The pace of data production has recently increased with the adoption of next-generation sequencing technologies and large-scale data production efforts to discover the breadth of genomic variation in humans. Comprehensive analysis of these datasets requires the coordinated use of Web-based applications and data repositories, desktop analysis tools and visualizers, and single-purpose algorithms. However, the effort required to transfer data between tools, convert between data formats, and manage results often prevents researchers from utilizing the wealth of methods available to them. Many integrative genomics and translational “bench to bedside” discoveries are possible with combinations of existing tools, but the necessary transitions between them puts them out of the reach of most researchers. GenomeSpace, http://www.genomespace.org, is an environment that brings together diverse computational tools, enabling scientists without programming skills to easily combine their capabilities. It aims to offer a common space to create, manipulate and share an ever-growing range of genomic analysis tools. GenomeSpace features support for cloud-based data storage and analysis, multi-tool analytic workflows, automatic conversion of data formats, and ease of connecting new tools to the environment. A set of six “GenomeSpace-enabled” seed tools developed by collaborating organizations provides a comprehensive platform for the analysis of cancer data: Cytoscape (UCSD), Galaxy (Penn State University), GenePattern (Broad Institute), Genomica (Weitzmann Institute), Integrative Genomics Viewer (Broad Institute), and the UCSC Genome Browser (UCSC). We show how researchers can use GenomeSpace to effortlessly combine the capabilities of all of these tools in several cancer research scenarios. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3966. doi:1538-7445.AM2012-3966

International Cancer Genome Consortium Data Portal--a one-stop shop for cancer genomics data

The International Cancer Genome Consortium (ICGC) is a collaborative effort to characterize genomic abnormalities in 50 different cancer types. To make this data available, the ICGC has created the ICGC Data Portal. Powered by the BioMart software, the Data Portal allows each ICGC member institution to manage and maintain its own databases locally, while seamlessly presenting all the data in a single access point for users. The Data Portal currently contains data from 24 cancer projects, including ICGC, The Cancer Genome Atlas (TCGA), Johns Hopkins University, and the Tumor Sequencing Project. It consists of 3478 genomes and 13 cancer types and subtypes. Available open access data types include simple somatic mutations, copy number alterations, structural rearrangements, gene expression, microRNAs, DNA methylation and exon junctions. Additionally, simple germline variations are available as controlled access data. The Data Portal uses a web-based graphical user interface (GUI) to offer researchers multiple ways to quickly and easily search and analyze the available data. The web interface can assist in constructing complicated queries across multiple data sets. Several application programming interfaces are also available for programmatic access. Here we describe the organization, functionality, and capabilities of the ICGC Data Portal.Database URL: http://dcc.icgc.org

Abstract 51: International Cancer Genome Consortium Data Portal – A ‘one-stop-shop’ for genomic, transcriptomic, and epigenomic data

Abstract The International Cancer Genome Consortium (ICGC) (www.icgc.org) orchestrates a multi-national effort to catalogue genomic abnormalities in 50 different tumor types. For each type 500 pairs of matched tumor and normal tissues will be studied using multiple technologies including DNA sequencing, expression profiling, and epigenetic analysis. ICGC data will be linked to cancer etiology, drug response, and patient survival. Each ICGC member institute generates its site-specific data independently and stores it in a local database. These databases are federated using BioMart technology (www.biomart.org) and made available to the public through the ICGC data portal (dcc.icgc.org). To the user, the multiple databases appear as a single integrated database. The ICGC data portal provides a platform for scientists to search, download, and analyze a broad range of cancer data. In addition to the data generated by the ICGC members, the portal includes data from other projects such as The Cancer Genome Atlas (TCGA) and Tumor Sequencing Project (TSP). Currently, the portal contains data sets from 15 cancer types. Data sets are divided into two categories: open access and controlled access. Open access data sets are available to the entire research community and include somatic mutations (simple, copy number variations, and structural rearrangements), gene and miRNA expression, DNA methylation, and exon junctions. Controlled access data sets, such as germline mutations, are available to users who have been authorized according to ICGC policies. The portal also includes annotation data from other publically available databases such as Ensembl, KEGG, Reactome, and the Pancreatic Expression Database. All data in the portal is exposed through a variety of user interfaces and can be processed using available analytical and visualization tools. For instance, users can perform gene- or pathway-centric analyses, such as compare mutation frequencies in specific genes or pathways, or compare mutation patterns between multiple tumor samples. Furthermore, by selecting diverse query and output criteria, users can retrieve data based on different molecular and clinical covariates. The portal enables users to conveniently retrieve, characterize, and compare cancer data from different sources through a single point of access. As more ICGC data is added and portal's functionality is further enhanced, it is expected that it will become an increasingly important resource for cancer researchers with diverse user requirements. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 51. doi:10.1158/1538-7445.AM2011-51

Abstract PL04-01: Improving patient outcomes in lung cancer by targeting different driver mutations

Abstract Multiple groups reported in 2004 on the association between clinical response to treatment with gefitinib and erlotinib and somatic mutations of the epidermal growth factor receptor (EGFR) in patients with lung cancer. Subsequent research showed these somatic sensitizing mutations of EGFR are also driver mutations. Somatic mutations of EGFR are predictive markers for higher response rates, prolonged time to progression, and a trend towards longer survival in patients with non-small cell lung cancer (NSCLC) treated with gefitinib or erlotinib rather than combination chemotherapy. The prospective clinical trials have led to approval of gefitinib for as initial treatment for patients with sensitizing mutations of EGFR or following relapse after conventional chemotherapy in Europe. Erlotinib is recommended as initial therapy for patients with sensitizing mutations of EGFR in the National Comprehensive Cancer Network (NCCN) guidelines. However, only 10% to 30% of patients with lung cancer have activating mutations of EGFR. Ongoing research is attempting to identify additional driver mutations in patients with lung cancer and other malignancies that can be targeted with existing or novel compounds. The activating mutations in both adenocarcinomas and squamous cell carcinomas of the lung are being systemically characterized through the Tumor Sequencing Project, The Cancer Genome Atlas projects (http://cancergenome.nih.gov/wwd/program), and other programs around the world. These projects have identified the most frequently mutated and/or activated genes in both lung adenocarcinoma and in squamous cell carcinoma of the lung. In addition to EGFR mutations, other driver mutations that can be effectively treated with existing targeted agents have been identified including EML4-ALK intrachromosomal rearrangements and BRAF mutations. Investigators from Japan discovered a gene that arose from a intrachromosomal rearrangement in adenocarcinomas of the lung which could transform NIH 3T3 cells. The transforming gene is a fusion of the ALK gene with echinoderm microtubule-associated protein-like 4 (EML4). The chromosomal rearrangement gives rise to a fusion gene in which the ALK tyrosine kinase is constitutively activated. Further studies have shown the EML4-ALK rearrangements are present in NSCLC arising in patients from the United States and Europe. The translocated gene can now be detected by using fluorescence in situ hybridization (FISH) in histological sections of the tumor. There are drugs that are directed against the ALK tyrosine kinase including PF2341066 or crizotinib. Crizotinib has been tested in patients with EML4-ALK intrachromosomal rearrangements in an expansion cohort of the phase I trials. Crizotinib has shown evidence of antitumor activity with response rates of approximately 57% and progression-free survival in excess of 6 months in patients with this rearrangement. Crizotinib is being tested in patients with relapsed NSCLC and EML4-ALK translocations randomized to either conventional therapy with pemetrexed or docetaxel versus crizotinib (ClinicalTrials.gov NCT01000025). Other potential therapeutic targets in patients with NSCLC include activating mutations in BRAF, present in 2% of patients with NSCLC. BRAF has been shown to be a driver mutations in melanoma and successfully targeted with PLX4032 {Bollag, #4046; Flaherty, #4045}. One thousand adenocarcinomas of the lung are now being characterized for 10 different mutations, gene amplification, and chromosomal rearrangements through the 14 institutions participating in the Lung Cancer Mutation Consortium (http://www.golcmc.com/). The Clinical Trials Committee within the Lung Cancer Mutation Consortium has established trials with drugs specifically directed at these driver mutations in different subsets of patients with lung cancer. The goal is to continue to expand the number of patients who can be treated with effective targeted therapy against these driver mutations. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr PL04-01. doi:10.1158/1538-7445.AM2011-PL04-01

PathScan: a tool for discerning mutational significance in groups of putative cancer genes

The expansion of cancer genome sequencing continues to stimulate development of analytical tools for inferring relationships between somatic changes and tumor development. Pathway associations are especially consequential, but existing algorithms are demonstrably inadequate. Here, we propose the PathScan significance test for the scenario where pathway mutations collectively contribute to tumor development. Its design addresses two aspects that established methods neglect. First, we account for variations in gene length and the consequent differences in their mutation probabilities under the standard null hypothesis of random mutation. The associated spike in computational effort is mitigated by accurate convolution-based approximation. Second, we combine individual probabilities into a multiple-sample value using Fisher-Lancaster theory, thereby improving differentiation between a few highly mutated genes and many genes having only a few mutations apiece. We investigate accuracy, computational effort and power, reporting acceptable performance for each. As an example calculation, we re-analyze KEGG-based lung adenocarcinoma pathway mutations from the Tumor Sequencing Project. Our test recapitulates the most significant pathways and finds that others for which the original test battery was inconclusive are not actually significant. It also identifies the focal adhesion pathway as being significantly mutated, a finding consistent with earlier studies. We also expand this analysis to other databases: Reactome, BioCarta, Pfam, PID and SMART, finding additional hits in ErbB and EPHA signaling pathways and regulation of telomerase. All have implications and plausible mechanistic roles in cancer. Finally, we discuss aspects of extending the method to integrate gene-specific background rates and other types of genetic anomalies. PathScan is implemented in Perl and is available from the Genome Institute at: http://genome.wustl.edu/software/pathscan.

Identifying cancer driver genes in tumor genome sequencing studies

Motivation: Major tumor sequencing projects have been conducted in the past few years to identify genes that contain ‘driver’ somatic mutations in tumor samples. These genes have been defined as those for which the non-silent mutation rate is significantly greater than a background mutation rate estimated from silent mutations. Several methods have been used for estimating the background mutation rate.Results: We propose a new method for identifying cancer driver genes, which we believe provides improved accuracy. The new method accounts for the functional impact of mutations on proteins, variation in background mutation rate among tumors and the redundancy of the genetic code. We reanalyzed sequence data for 623 candidate genes in 188 non-small cell lung tumors using the new method. We found several important genes like PTEN, which were not deemed significant by the previous method. At the same time, we determined that some genes previously reported as drivers were not significant by the new analysis because mutations in these genes occurred mainly in tumors with large background mutation rates.Availability: The software is available at: http://linus.nci.nih.gov/Data/YounA/software.zipContact: rsimon@mail.nih.govSupplementary information: Supplementary data are available at Bioinformatics online.