A position in Paris funded specifically for non-French scientists (or French working abroad for some time). Post-doc position: Regularization methods for genomic selection Job description In genomic selection, one aims at predicting the performance of individuals according to their available genotypic information (molecular markers). To handle the computational problems arising when the number of descriptors (markers) is higher than the number of observations (individuals), classical regularized methods have been applied. These methods could be improved to take the distinctive features of the data into account through biological priors, to enhance both the performance and interpretability of the prediction function. In this context, the post-doc fellow's research will focus on the following two important issues: 1. the development of new regularization functions to handle biological prior knowledge available for genetic data, i.e. linkage disequilibrium and gene density distribution along the genome, 2. the development of "multi-task" regularization functions for the joint prediction of a given phenotype in several populations. Desired profile The targeted candidate should have a good experience in large dataset analysis. The ideal profile includes: - a strong knowledge in statistics for high-dimensional data analysis, - a good knowledge of R and/or object-oriented programming (C++), - an experience in optimization and/or computational statistics. The applicant should be competent, highly motivated, interactive and should have an interest for applications in genetics. Context This is a 2-year position, located at AgroParisTech in Paris, France (http://www.agroparistech.<http://www.agroparistech.fr/mia/>fr/mia/<http://www.agroparistech.fr/mia/> ), in the Statistics and Genomics team headed by Stephane Robin. Applications to animal and plant models will be performed in collaboration with Alain Charcosset (plant genetics, INRA) and Vincent Ducrocq (animal genetics, INRA). The position is eligible to the Agreenskills excellence programme for young researchers, offering attractive conditions in terms of salary. Contact Candidates should send a detailed CV with a cover letter and references to Julien Chiquet (julien.<mailto:julien.chiquet@genopole.cnrs.fr>chiquet@genopole.cnrs.fr<mailto:julien.chiquet@genopole.cnrs.fr>) and Tristan Mary-Huard (maryhuar@agroparistech.fr<mailto:maryhuar@agroparistech.fr> ). References Sparsity in sign-coherent groups of variables via the cooperative-Lasso, J.Chiquet, Y. Grandvalet, and C. Charbonnier. Annals of Applied Statistics, 2012. Accounting for linkage disequilibrium in regularized methods for multi-trait genomic selection, J. Chiquet, T. Mary-Huard, S. Robin. Submitted. Regularization Methods for High-Dimensional Instrumental Variables Regression With an Application to Genetical Genomics, W. Lin, R. Feng, and H. Li, http://arxiv-web3.library.cornell.<http://arxiv-web3.library.cornell.edu/abs/1304.7829>edu/abs/1304.7829<http://arxiv-web3.library.cornell.edu/abs/1304.7829>