GPI Seminar Series: Patrick Emmanuel Meyer

Patrick Emmanuel MeyerBiosys Unit (ULg, Belgium), Biological Network Inference, from dependency to causality and implication.
Tuesday October 28th, 10h, Sala de Seminaris D5-007

Biological Network Inference, from dependency to causality and implication

Abstract:
In this presentation, we introduce the various concepts of causality and information used behind biological (expression-based transcriptional) network inference. We show that bivariate (or pairwise) information measures such as mutual information or correlations are heavily used in those network inference algorithms mainly because they are faster and require fewer samples than multivariate (or multidimensional) strategies. We also introduce a new bivariate measure, called the rank minrelation coefficient, that improves the selection of relevant genes, in particular when compared to correlations and mutual information.

Patrick's short bio: 
Patrick Emmanuel Meyer received the Electromechanical Engineering degree in 2003 and the Ph.D. degree in Sciences (statistical machine learning), in 2008, from the Université Libre de Bruxelles (ULB, Belgium). After postdoctoral research at the Computer Science and Artificial Intelligence Laboratory of the Massachusetts Institute of Technology (CSAIL, MIT, USA, 2010), at the BROAD Institute of MIT and Harvard (Boston, USA, 2011) and at the FNRS (Belgium, 2012), he became, in 2014, Professor in Bioinformatics and Systems Biology at the Université de Liège (ULg, Belgium).

Among other scientific productions, he co-authored, the Drosophila modENCODE paper published in Science and the open-source R and Bioconductor package, Mutual Information NETworks. His interests cover statistical machine learning, information theory and systems biology.