Dec 14, 2011
from 02:30 PM to 03:30 PM
|Where||Keynes Hall, King's College|
|Contact Name||Petra Vertes|
|Add event to calendar||
Francesco Iorio (European Bioinformatics Institute & Sanger Institute):
Drug Discovery and Re-purposing by Network-Analysis of Gene Expression Data
A significant number of recent studies wink at the idea that every biological state can be described by a proper gene expression signature: a well defined set of genes together with a pattern of expression that is exclusively linked to it. The underlying concepts are: i) any condition, for example, the activity of a given pathway, a disease phenotype or cellular response to drug treatment, realizes some change in transcriptional activity; ii) even if a single gene on its own poorly characterizes a biological state of interest, this ability is significantly improved when considering a set of genes with their combined pattern of expression.
I will describe MANTRA (Mode of Action by NeTwoRk Analysis): a computational tool for the analysis of the Mode of Action (MoA) of novel drugs and the identification of known and approved candidates for "drug repositioning", combining the introduced ideas with simple concepts from network-theory and non-parametric statistics.
Florian Markowetz (Cancer Research UK Cambridge Research Institute):
Inferring phenotypic networks with Nested Effect Models
In high-dimensional phenotyping screens, a large number of cellular features is observed after perturbing genes by knockouts or RNA interference. Comprehensive analysis of perturbation effects is one of the most powerful techniques for attributing functions to genes. I will talk about Nested Effects Models, a probabilistic method to efficiently infer a genetic hierarchy from the nested structure of observed perturbation effects. These hierarchies elucidate the structures of signaling pathways and regulatory networks. Our methods achieve two goals: (1) they reveal clusters of genes with highly similar phenotypic profiles, and (2) they order (clusters of) genes according to subset relationships between phenotypes. In contrast to other graphical models, Nested Effect Models are not built on conditional independence, but subset relations - making them attractive from a theoretical and applied perspective.