Discovering Haplotypes in Linkage Disequilibrium Mapping with an Adaptive Genetic Algorithm

In this paper, we present an evolutionary approach to discover candidate haplotypes in a linkage disequilibrium study. This work takes place into the study of factors involved in multi-factorial diseases such as diabetes and obesity. A first study on the linkage disequilibrium problem structure led us to use a genetic algorithm to solve it. Due to the particular, but classical, evaluation function given by the biologists, we design our genetic algorithm with several populations. This model lead us to implement different cooperative operators such as mutation and crossover. Probabilities of application of those mechanisms are set adaptively. In order to introduce some diversity, we also implement a random immigrant strategy and to cover up the cost of the evaluation computation we parallelize it in a master / slave model. Different combinations of the presented mechanisms are tested on real data and compared in term of robustness and computation cost. We show that the most complete strategy is able to find the best solutions and is the most robust.