The cancer stem cell hypothesis, that a small population of tumour cells are responsible for tumorigenesis and cancer progression, is becoming widely accepted and recent evidence has suggested a prognostic and predictive role for such cells. Intra-tumour heterogeneity, the diversity of the cancer cell population within the tumour of an individual patient, is related to cancer stem cells and is also considered a potential prognostic indicator in oncology. The measurement of cancer stem cell abundance and intra-tumour heterogeneity in a clinically relevant manner however, currently presents a challenge. Here we propose signalling entropy, a measure of signalling pathway promiscuity derived from a sample's genome-wide gene expression profile, as an estimate of the stemness of a tumour sample. By considering over 500 mixtures of diverse cellular expression profiles, we reveal that signalling entropy also associates with intra-tumour heterogeneity. By analysing 3668 breast cancer and 1692 lung adenocarcinoma samples, we further demonstrate that signalling entropy correlates negatively with survival, outperforming leading clinical gene expression based prognostic tools. Signalling entropy is found to be a general prognostic measure, valid in different breast cancer clinical subgroups, as well as within stage I lung adenocarcinoma. We find that its prognostic power is driven by genes involved in cancer stem cells and treatment resistance. In summary, by approximating both stemness and intra-tumour heterogeneity, signalling entropy provides a powerful prognostic measure across different epithelial cancers.

BACKGROUND: The use of tissue microarrays (TMA) and advances in digital scanning microscopy has enabled the collection of thousands of tissue images. There is a need for software tools to annotate, query and share this data amongst researchers in different physical locations. RESULTS: We have developed an open source web-based application for remote scoring of TMA images, which exploits the value of Microsoft Silverlight Deep Zoom to provide a intuitive interface for zooming and panning around digital images. We use and extend existing XML-based standards to ensure that the data collected can be archived and that our system is interoperable with other standards-compliant systems. CONCLUSION: The application has been used for multi-centre scoring of TMA slides composed of tissues from several Phase III breast cancer trials and ten different studies participating in the International Breast Cancer Association Consortium (BCAC). The system has enabled researchers to simultaneously score large collections of TMA and export the standardised data to integrate with pathological and clinical outcome data, thereby facilitating biomarker discovery.

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.

MOTIVATION: In the past decade, a number of technologies to quantify allele-specific expression (ASE) in a genome-wide manner have become available to researchers. We investigate the application of single-nucleotide polymorphism (SNP) microarrays to this task, exploring data obtained from both cell lines and primary tissue for which both RNA and DNA profiles are available. RESULTS: We analyze data from two experiments that make use of high-density Illumina Infinium II genotyping arrays to measure ASE. We first preprocess each data set, which involves removal of outlier samples, careful normalization and a two-step filtering procedure to remove SNPs that show no evidence of expression in the samples being analyzed and calls that are clear genotyping errors. We then compare three different tests for detecting ASE, one of which has been previously published and two novel approaches. These tests vary at the level at which they operate (per SNP per individual or per SNP) and in the input data they require. Using SNPs from imprinted genes as true positives for ASE, we observe varying sensitivity for the different testing procedures that improves with increasing sample size. Methods that rely on RNA signal alone were found to perform best across a range of metrics. The top ranked SNPs recovered by all methods appear to be reasonable candidates for ASE. AVAILABILITY AND IMPLEMENTATION: Analysis was carried out in R (http://www.R-project.org/) using existing functions.

UNLABELLED: Copy number aberrations are recognized to be important in cancer as they may localize to regions harboring oncogenes or tumor suppressors. Such genomic alterations mediate phenotypic changes through their impact on expression. Both cis- and transacting alterations are important since they may help to elucidate putative cancer genes. However, amidst numerous passenger genes, trans-effects are less well studied due to the computational difficulty in detecting weak and sparse signals in the data, and yet may influence multiple genes on a global scale. We propose an integrative approach to learn a sparse interaction network of DNA copy-number regions with their downstream transcriptional targets in breast cancer. With respect to goodness of fit on both simulated and real data, the performance of sparse network inference is no worse than other state-of-the-art models but with the advantage of simultaneous feature selection and efficiency. The DNA-RNA interaction network helps to distinguish copy-number driven expression alterations from those that are copy-number independent. Further, our approach yields a quantitative copy-number dependency score, which distinguishes cis- versus trans-effects. When applied to a breast cancer data set, numerous expression profiles were impacted by cis-acting copy-number alterations, including several known oncogenes such as GRB7, ERBB2, and LSM1. Several trans-acting alterations were also identified, impacting genes such as ADAM2 and BAGE, which warrant further investigation. AVAILABILITY: An R package named lol is available from www.markowetzlab.org/software/lol.html.

The influence of DNA cis-regulatory elements on a gene's expression has been intensively studied. However, little is known about expressions driven by trans-acting DNA hotspots. DNA hotspots harboring copy number aberrations are recognized to be important in cancer as they influence multiple genes on a global scale. The challenge in detecting trans-effects is mainly due to the computational difficulty in detecting weak and sparse trans-acting signals amidst co-occuring passenger events. We propose an integrative approach to learn a sparse interaction network of DNA copy-number regions with their downstream targets in a breast cancer dataset. Information from this network helps distinguish copy-number driven from copy-number independent expression changes on a global scale. Our result further delineates cis- and trans-effects in a breast cancer dataset, for which important oncogenes such as ESR1 and ERBB2 appear to be highly copy-number dependent. Further, our model is shown to be efficient and in terms of goodness of fit no worse than other state-of the art predictors and network reconstruction models using both simulated and real data.

The CancerGrid consortium is developing open- standards cancer informatics to address the challenges posed by modern cancer clinical trials. This paper presents the service-oriented software paradigm implemented in CancerGrid to derive clinical trial information management systems for collaborative cancer research across multiple institutions. Our proposal is founded on a combination of a clinical trial (meta)model and WSRF (Web Services Resource Framework), and is currently being evaluated for use in early phase trials. Although primarily targeted at cancer research, our approach is readily applicable to other areas for which a similar information model is available.

Cancer clinical trials pose significant challenges to the e-Science community. The information technology required to enable this kind of large-scale, collaborative science will need to support easy and rapid development and deployment of reliable and flexible software systems that enable syntactic, semantic and computational interoperability. CancerGrid, an e-Science consortium funded by the UK Medical Research Council, is addressing these challenges through the development of model-driven, service-oriented technology for cancer informatics. This poster presents recent significant efforts in CancerGrid, resulting in the metamodel-based automated generation of WSRF (Web Services Resource Framework) compliant trial management systems. The most important advantages of our approach are discussed.

The quantity of mRNA transcripts in a cell is determined by a complex interplay of cooperative and counteracting biological processes. Independent Component Analysis (ICA) is one of a few number of unsupervised algorithms that have been applied to microarray gene expression data in an attempt to understand phenotype differences in terms of changes in the activation/inhibition patterns of biological pathways. While the ICA model has been shown to outperform other linear representations of the data such as Principal Components Analysis (PCA), a validation using explicit pathway and regulatory element information has not yet been performed. We apply a range of popular ICA algorithms to six of the largest microarray cancer datasets and use pathway-knowledge and regulatory-element databases for validation. We show that ICA outperforms PCA and clustering-based methods in that ICA components map closer to known cancer-related pathways, regulatory modules, and cancer phenotypes. Furthermore, we identify cancer signalling and oncogenic pathways and regulatory modules that play a prominent role in breast cancer and relate the differential activation patterns of these to breast cancer phenotypes. Importantly, we find novel associations linking immune response and epithelial-mesenchymal transition pathways with estrogen receptor status and histological grade, respectively. In addition, we find associations linking the activity levels of biological pathways and transcription factors (NF1 and NFAT) with clinical outcome in breast cancer. ICA provides a framework for a more biologically relevant interpretation of genomewide transcriptomic data. Adopting ICA as the analysis tool of choice will help understand the phenotype-pathway relationship and thus help elucidate the molecular taxonomy of heterogeneous cancers and of other complex genetic diseases.

MOTIVATION: Elucidating the molecular taxonomy of cancers and finding biological and clinical markers from microarray experiments is problematic due to the large number of variables being measured. Feature selection methods that can identify relevant classifiers or that can remove likely false positives prior to supervised analysis are therefore desirable. RESULTS: We present a novel feature selection procedure based on a mixture model and a non-gaussianity measure of a gene's expression profile. The method can be used to find genes that define either small outlier subgroups or major subdivisions, depending on the sign of kurtosis. The method can also be used as a filtering step, prior to supervised analysis, in order to reduce the false discovery rate. We validate our methodology using six independent datasets by rediscovering major classifiers in ER negative and ER positive breast cancer and in prostate cancer. Furthermore, our method finds two novel subtypes within the basal subgroup of ER negative breast tumours, associated with apoptotic and immune response functions respectively, and with statistically different clinical outcome. AVAILABILITY: An R-function pack that implements the methods used here has been added to vabayelMix, available from (www.cran.r-project.org). CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary information is available at Bioinformatics online.

MOTIVATION: Accurate subcategorization of tumour types through gene-expression profiling requires analytical techniques that estimate the number of categories or clusters rigorously and reliably. Parametric mixture modelling provides a natural setting to address this problem. RESULTS: We compare a criterion for model selection that is derived from a variational Bayesian framework with a popular alternative based on the Bayesian information criterion. Using simulated data, we show that the variational Bayesian method is more accurate in finding the true number of clusters in situations that are relevant to current and future microarray studies. We also compare the two criteria using freely available tumour microarray datasets and show that the variational Bayesian method is more sensitive to capturing biologically relevant structure.