Cladistic analysis of haplotypes as an attempt to detect disease susceptibility

We report the results of our analyses of the Genetic Analysis Workshop 12 s imulated data set, using phylogenetic methods to reconstruct the history of haplotypes. We selected candidate gene I and 6, drawn from the isolate pop ulation. In a first step, haplotypes were inferred using family data. In a second step, cladistic approaches were performed to select the most parsimo nious trees under various conditions of character state transformation and ancestral hypotheses, in order to check whether the affected status is more frequent in some clades than in others. Sites which are synapormophies of such clades can be viewed as candidate sites for the disease susceptibility . The method seems to be efficient for the candidate gene 1, but not for th e candidate gene 6. Effects of the genetic model underlying the affection s tatus are discussed, particularly dominance, penetrance, and, in the contex t of this simulated data, procedure followed to generate haplotypes. These preliminary results deserve further investigations. (C) 2001 Wiley-Liss, In c.