Genetic variants underlying complex traits, including disease susceptibility, are enriched within the transcriptional regulatory elements, promoters and enhancers. There is emerging evidence that regulatory elements associated with particular traits or diseases share similar patterns of transcriptional activity. Accordingly, shared transcriptional activity (coexpression) may help prioritise loci associated with a given trait, and help to identify underlying biological processes. Using cap analysis of gene expression (CAGE) profiles of promoter- and enhancer-derived RNAs across 1824 human samples, we have analysed coexpression of RNAs originating from trait-associated regulatory regions using a novel quantitative method (network density analysis; NDA). For most traits studied, phenotype-associated variants in regulatory regions were linked to tightly-coexpressed networks that are likely to share important functional characteristics. Coexpression provides a new signal, independent of phenotype association, to enable fine mapping of causative variants. The NDA coexpression approach identifies new genetic variants associated with specific traits, including an association between the regulation of the OCT1 cation transporter and genetic variants underlying circulating cholesterol levels. NDA strongly implicates particular cell types and tissues in disease pathogenesis. For example, distinct groupings of disease-associated regulatory regions implicate two distinct biological processes in the pathogenesis of ulcerative colitis; a further two separate processes are implicated in Crohn's disease. Thus, our functional analysis of genetic predisposition to disease defines new distinct disease endotypes. We predict that patients with a preponderance of susceptibility variants in each group are likely to respond differently to pharmacological therapy. Together, these findings enable a deeper biological understanding of the causal basis of complex traits.

MOTIVATION: Quantification of the contribution of genetic variation to phenotypic variation for complex traits becomes increasingly computationally demanding with increasing numbers of single-nucleotide polymorphisms and individuals. To meet the challenges in making feasible large-scale studies, we present the REgional heritability advanced complex trait analysis software. Adapted from advanced complex trait analysis (and, in turn, genome-wide complex trait analysis), it is tailored to exploit the parallelism present in modern traditional and graphics processing unit (GPU)-accelerated machines, from workstations to supercomputers. RESULTS: We adapt the genetic relationship matrix estimation algorithm to remove limitations on memory, allowing the analysis of large datasets. We build on this to develop a version of the code able to efficiently exploit GPU-accelerated systems for both the genetic relationship matrix and REstricted maximum likelihood (REML) parts of the analysis, offering substantial speedup over the traditional central processing unit version. We develop the ability to analyze multiple small regions of the genome across multiple compute nodes in parallel, following the 'regional heritability' approach. We demonstrate the new software using 1024 GPUs in parallel on one of the world's fastest supercomputers. AVAILABILITY: The code is freely available at http://www.epcc.ed.ac.uk/software-products CONTACT: [email protected].

MOTIVATION: The Genome-wide Complex Trait Analysis (GCTA) software package can quantify the contribution of genetic variation to phenotypic variation for complex traits. However, as those datasets of interest continue to increase in size, GCTA becomes increasingly computationally prohibitive. We present an adapted version, Advanced Complex Trait Analysis (ACTA), demonstrating dramatically improved performance. RESULTS: We restructure the genetic relationship matrix (GRM) estimation phase of the code and introduce the highly optimized parallel Basic Linear Algebra Subprograms (BLAS) library combined with manual parallelization and optimization. We introduce the Linear Algebra PACKage (LAPACK) library into the restricted maximum likelihood (REML) analysis stage. For a test case with 8999 individuals and 279,435 single nucleotide polymorphisms (SNPs), we reduce the total runtime, using a compute node with two multi-core Intel Nehalem CPUs, from ∼17 h to ∼11 min. AVAILABILITY AND IMPLEMENTATION: The source code is fully available under the GNU Public License, along with Linux binaries. For more information see http://www.epcc.ed.ac.uk/software-products/acta. CONTACT: [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Blue Gene was designed as a special purpose system, offering unprecedented computing performance coupled with very low power consumption and cost for a limited set of applications. Adoption has however become more widespread and Blue Gene is now being used in a relatively diverse range of scientific disciplines. Exploring the range and type of application that can make effective use of these systems, and the relevant architectural features influencing performance, is therefore highly topical.This tutorial will focus on the performance of a series of applications and consider the techniques required to achieve optimal performance and scaling on the Blue Gene architecture. We will discuss the factors and bottlenecks that influence performance and cover techniques for compiler, memory and communication optimization.Attendees will have the opportunity to gain hands on experience through a series of practical sessions on The University of Edinburgh's Blue Gene system.

This paper briefly describes the picoArray/spl trade/ architecture, and in particular the deterministic internal communication fabric. The methods that have been developed for debugging and verifying systems using devices from the picoArray family are explained. In order to maximize the computational ability of these devices, hardware debugging support has been kept to a minimum and methods and tools developed to take this into account.

Searching and organization of peers are fundamental challenges in P2P networks. Unstructured networks, such as Gnutella, inefficiently use broadcast searches and random neighbors. Structured networks are similarly inefficient, as they generally rely on globally unique identifiers (GUIDs) which are assigned irrespective of content, which prevents fuzzy semantic searches. In both types of network search, neighbors establish trust between themselves, regardless of whether or not their content is likely to satisfy searches. We present the idea of using context-based profiles to describe peers. This enables self-organizing clusters of similar peers. A profile represents a peer's expertise based on content and responsiveness. By refining the search process using these profiles, more efficient directed searches are possible. Moreover, expertise provides a basis for trust establishment.