Statistical methods for omics analysis

We develop original statistical methods to analyze high-throughput molecular biology data. Our goal is to cope with the intrinsic difficult of analyzing such data set, in particular their high dimensionality, while facilitating their biological interpretation. We are experts of model-based approach for clustering, linear models and algorithms for molecular biology. These methods allowed us to investigate genome activity: gene expression, gene interactions and control mechanisms. Main topics are about:

• Neutral comparison studies
• Development of mixture models for co-expression networks and protein-DNA interactions
• Segmentation methods for longitudinal analysis of the genome.
• Statistical inference of gene networks.