Biomedical research, be it in the biopharmaceutical industry, academia or governmental laboratories, is becoming an information driven science. An enormous amount of data is either present in published literature and patents or being generated using high-throughput technologies (such as DNA sequencing, microarrays, genome-wide association, proteomics etc.). The resulting computational biology challenges currently involve the development of intelligent data mining methodologies for these high-dimensional data and the development of databases that facilitate the hosting and integration of these diverse data-sets. The computational advances being made in these areas are significantly impacting biomedical research ranging from fundamental biological findings that relate genes and environment to disease all the way to molecular biomarkers for drug efficacy or toxicity.
In our InnovationWell session chaired by Debraj GuhaThakurta (Rosetta Inpharmatics, Merck & Co.) taking place 14 October 2008 at Bryn Mawr, we have gathered a set of top researchers who have significant experience in the areas of data mining and integration. Presentation and discussion topics will include: molecular network reconstruction from genetically segregating populations (Paul McDonagh), the use of molecular networks for environmental risk assessment (Stephen Edwards), semantic web technologies for modeling biological pathways (Christopher Bouton), application of integrated genomics and genetics in pharmaceutical discovery (Debraj GuhaThakurta), and comparative genomics in drug discovery (James Brown).
A description of the session with presentation abstracts follows. Please add your comments, discussion or questions at the end of the post.
Computational Biology
http://innovationwell.net/comty_confprogr08compbio
(Please follow continuation here to read abstracts. Comments can be made at the end.)
Abstracts
Comparative Genomics and Drug Discovery
James R. Brown, M.Sc., Ph.D. Team Leader, Infectious Disease and External Drug Discovery Computational Biology, GlaxoSmithKline, 1250 South Collegeville Road, UP1345, P.O. Box 5089, Collegeville PA 19426-0989
The rapid growth of genomic sequence data provides unparalleled opportunities for comparative genomics and evolutionary biology research. Comparative genomics and molecular evolutionary approaches are seeing increasing applications in pharmaceutical research, making a substantial contribution to our understanding of disease and drug targets. The development of new antibiotics and understanding resistance to existing drugs is one area where both intra- and interspecific comparative genomics is essential. However, the availability of genomes from a growing number of mammalian species as well as from multiple human individuals is also impacting other therapeutic areas, such as cancer. Acknowledging the mosaic nature of the human genome, being comprised of genes with widely different evolutionary trajectories, is an important new paradigm for drug discovery.
Semantic Web Modeling of Biological Systems
Christopher M.L.S. Bouton, Ph.D., Head of Integrative Data Mining & Computational Biology Group Leader, Research Technology Center, Pfizer Inc.
Semantic web technologies provide the framework for the development of conceptual networks and the inferential reasoning through these networks. We explore the use of these technologies in the modeling and analysis of biological pathways.
Systems Biology and Mode of Action Based Risk Assessment
Stephen W. Edwards, Ph.D. , Principal Investigator of Systems Biology, ADHIO, NHEERL, ORD, U.S. Environmental Protection Agency (B305-01), 109 TW Alexander Drive, Research Triangle Park, NC 27711, USA
The application of systems biology has increased in the past decade largely as a consequence of the human genome project and technological advances in genomics and proteomics. Systems approaches have been used in the medical & pharmaceutical realm for diagnostic purposes and target identification. During this same period, risk assessment has also been transformed by a variety of factors including a much greater emphasis on mode of action (MOA) in defining risk. MOA is defined as “a brief description of the sequence of measured events” from chemical administration to adverse outcome. Genome-wide measurements provide both a discovery engine for identifying the MOA and a framework for evaluation of the MOA during the conduct of a risk assessment. This framework is important as these measurements are not chosen based on the hypothesized MOA and therefore represent an unbiased check of the comprehensiveness of the MOA. In addition, optimal design for MOA studies is critical to provide the time and dose dependent data required for quantitative model building. Finally, identification of biomarkers and bioindicators of disease in humans provides a viable way to extrapolate from disease outcomes measured at high exposure levels to those at low exposure levels and thus provide the opportunity to reduce or perhaps eliminate in vivo animal testing. To realize the full potential of these approaches, larger integrated projects which include all these individual components are necessary. [This abstract does not reflect EPA policy.]
Using Genetics, Gene Expression and Pathway Mining in Pharmaceutical Discovery: Application towards Metabolic and Cardiovascular Disorders
Debraj GuhaThakurta, Rosetta Inpharmatics, Merck & Co.
We are using an integrated genetics and genomics approach on preclinical and clinical samples to identify novel targets and biomarkers for complex diseases such as obesity, diabetes, atherosclerosis and hypertension. We have used mouse crosses as well as human cohorts to perform integrated analysis of phenotype, genotype and gene expression profiles for our discovery purposes. Regulatory networks have also been created from these data which inform us about the causal pathways strongly associated with diseases or potential disease biomarkers. Results from some of these studies will be presented along with an overview of how such findings are being integrated with additional orthogonal data to identify or prioritize new drug targets and candidate biomarkers.
Reverse Engineering and Simulating Causal Genetical Genomics Networks
Paul McDonagh Ph.D., Gene Network Sciences, 10 Canal Park, Cambridge, MA 02141, USA
Genetic and environmental variations cause phenotypic variations in animal populations. Experimental designs that harness or account for genetic variation can effectively uncover causal disease mechanisms in whole animals and will have a more relevant context than in vitro experiments with targeted gene disruptions. We show how to reconstruct an industrial-scale, in vivo genetical genomics model that exploits genetic variation and simultaneously includes thousands of genetic markers, liver gene expression data and lipid-related clinical variables from a mouse F2 intercross population using Bayesian network reconstruction techniques without the use of any constraining, prior information. We employ massively parallel machines to reverse engineer population of models that predict the ‘gearing’ between genetic and genomic molecular components with respect to lipid metabolism. Since a single topology or model cannot capture the reality of a biological system under investigation, Monte Carlo simulation strategies over the population of models reveal how the genetic, gene expression and clinical variables work together to impact the lipid-related clinical variables. The predictive power of our genetical genomics model can be demonstrated using Monte Carlo simulation to accurately predict relationships between clinical traits of mice that were not part of the F2 generation trained on. Monte-Carlo simulations also predict important causal genes in mouse lipid metabolism that would not otherwise be identified from either genetic or gene expression analysis alone.
Contact Information:
Program: Dr. Barry Hardy, InnovationWell Community of Practice & Research, Douglas Connect. Tel: +41 61 851 0170. barry.hardy -[at]- douglasconnect.com
Registration Enquiries: Nicki Douglas, Douglas Connect, Baermeggenweg 14, 4314 Zeiningen, Switzerland. Tel: +41 61 851 0461. InnovationWell -[at]- douglasconnect.com
or please visit:
http://InnovationWell-BM810.eventsbot.com
Barry Hardy
Yes this is promising new field; interestingly, it was foreshadowed in the book THE EPIC HISTORY OF BIOLOGY (basic books, publisher) by Anthony Serafini. I recommend this book for an excellent discussion of computational biology and genomic sequencing
Posted by: hobart | August 27, 2008 at 06:17 PM