Section: New Results

Regulation and signaling: detecting complex and discriminant signatures of phenotypes

Participants : Catherine Belleannée, Célia Biane-Fourati, Samuel Blanquart, Olivier Dameron, Maxime Folschette, Nicolas Guillaudeux, Marine Louarn, François Moreews, Anne Siegel, Nathalie Théret, Pierre Vignet, Méline Wéry.

Creation of predictive functional signaling networks [M. Folschette, N. Théret] [16].

• Integrating genome-wide gene expression patient profiles with regulatory knowledge is a challenging task because of the inherent heterogeneity, noise and incompleteness of biological data. We proposed an automatic pipeline to extract automatically regulatory knowledge from pathway databases and generate novel computational predictions related to the state of expression or activity of biological molecules. We applied it in the context of hepatocellular carcinoma (HCC) progression, and evaluated the precision and the stability of these computational predictions. Our computational model predicted the shifts of expression of 146 initially non-observed biological components. Our predictions were validated at 88% using a larger experimental dataset and cross-validation techniques.

Experimental evidences of transcript predictions [C. Belleannée, S. Blanquart, N. Guillaudeux] [13].

• We designed comparative-genomics based models of gene structures through genes comparisons across species. These models enable to predict putative transcript isoforms in a species given the knowledge available in other species [46]. We recently published a first experimental validation of such a predicted transcriptome [13]. In this work, transcript isoforms of the human TRPM8 gene yield transcript predictions in the mouse TRPM8 gene, which are experimentally validated using targeted PCR in mouse tissues. This work also provides a first attempt to estimate origin of new isoforms during the gene evolution.

• In another collaboration with IGDR, we considered a multi-species gene comparison including human, mouse and dog [39]. This work reveals global trends of the gene isoform sets evolution, suggesting a extremely high plasticity of alternative transcription and alternative splicing propensities in those three species. This work moreover provides experimental evidences of the predicted transcripts based on public RNAseq data, highlighting the tissue specificity of isoform expression across species.

Formalizing and enriching phenotype signatures using Boolean networks [C. Biane-Fourati, M. Wéry, A. Siegel, O. Dameron] [20], [42], [22]

• We used Formal Concept Analysis as a symbolic bi-clustering techniques to classify and sort the steady states of a Boolean network according to biological signatures based on the hierarchy of the roles the network components play in the phenotypes. We applied our approach on a T helper lymphocyte (Th) differentiation network with a set of signatures corresponding to the sub-types of Th. This led to the identification and prediction of a new hybrid sub-type later confirmed by the literature.