Please activate JavaScript!
Please install Adobe Flash Player, click here for download


76 ISCB 2014 Vienna, Austria • Abstracts - Oral PresentationsWednesday, 27th August 2014 • 14:00-15:30 Monday25thAugustTuesday26thAugustThursday28thAugustAuthorIndexPostersWednesday27thAugustSunday24thAugust be frequently mutated, with high or moderate mutational impact, exhibit- ing an extreme expression and functionally linked to a large number of differentially expressed neighbors in the gene network. We show that breast cancer patients carrying more mutated driver genes with functional implications and extreme expression pattern have worse survival than those with less mutated driver genes. We propose the driver gene score as an informative tool to predict survival for guiding patient care and clinical management.   C42.5 Nonparametric mixture modelling of dynamic Bayesian networks derives the structure of protein-networks in adhesion sites Y Fermin1 , K Ickstadt1 , MR Sheriff2 , S Imtiaz2 , E Zamir2 1 TU Dortmund University, Faculty of Statistics, Dortmund, Germany, 2 Max Planck Institute of Molecular Physiology, Dortmund, Germany   Cell-matrix adhesions play essential roles in important biological pro- cesses including cell morphogenesis, migration, proliferation, survival and differentiation (Gumbiner, 1996; Hynes and Lander, 1992).The attachment of cells to the extracellular matrix is mediated by dynamic sites along the plasma membrane, such as focal adhesions, at which receptors of the in- tegrin family anchor the actin cytoskeleton to components of the extra- cellular matrix via a large number of different proteins (Zamir and Geiger, 2001). Focal adhesions can contain over 100 different proteins, including integrins, adapter proteins, and intracellular signaling proteins (Zaidel-Bar et al., 2007). Due to the large number of components and diversity of cell- matrix adhesion sites, a fundamental question is how these sites are as- sembled and function. In systems biology graphical models and networks have been widely ap- plied as a useful tool to model complex biochemical systems. In this work we propose a nonparametric mixture of dynamic Bayesian networks to study interactions among proteins in the presence of the temporal struc- ture and heterogeneity among focal adhesions. Nonparametric mixture modelling of dynamic Bayesian networks is developed by a combination of dynamic Bayesian networks (Ghahramani, 1997) and of nonparametric Bayesian networks (Ickstadt et al., 2011). This approach provides further grouping of focal adhesions according to their network structures. We apply and illustrate our approach using multicolor live cell imaging datasets, in which the levels of four different proteins are monitored in individual focal adhesions. Keywords: Cell-matrix adhesions; Dynamic Bayesian networks; Nonparametric Bayesian networks  

Pages Overview