Learning a forest of Hierarchical Bayesian Networks to model dependencies between genetic markers

Raphaël Mourad (1) , Christine Sinoquet (1) , Philippe Leray (1)

1 LINA - Laboratoire d'Informatique de Nantes Atlantique

We propose a novel probabilistic graphical model dedicated to represent the statistical dependencies between genetic markers, in the Human genome. Our proposal relies on building a forest of hierarchical latent class models. It is able to account for both local and higher-order dependencies between markers. Our motivation is to reduce the dimension of the data to be further submitted to statistical association tests with respect to diseased/non diseased status. A generic algorithm, CFHLC, has been designed to tackle the learning of both forest structure and probability distributions. A first implementation of CFHLC has been shown to be tractable on benchmarks describing 100000 variables for 2000 individuals, on a standard personal computer.

Keywords: bioinformatics; biomedical data analysis; bayesian networks; hierarchical latent class model; data dimensionality reduction; modelling of genetic marker dependencies

Domains: Bioinformatics [q-bio.QM]; Quantitative Methods [q-bio.QM]

Raphaël Mourad, Christine Sinoquet, Philippe Leray. Learning a forest of Hierarchical Bayesian Networks to model dependencies between genetic markers. 2010. ⟨hal-00444087v2⟩