Date |
July 7, 2008 |
Speaker |
Dr. Jean-Philippe Vert, Ecole des Mines de Paris |
Title |
Inference of biological networks with supervised machine learning
|
Abstract |
In this talk I consider the problem of inferring edges in a biological network given some high-throughput genomic or proteomic data about the genes and proteins, which form the vertices of the network. While many "de novo" approaches attempt to directly infer the whole graph starting from nothing but the genomic data, I will discuss a series of "supervised" methods that use not only the genomic data but also the set of currently known edges in infer missing edges. Results on the reconstruction of the metabolic, protein-protein interaction and regulatory networks of the yeast and the bacteria E. coli will be presented.
|
|