[Linkage analysis for complex diseases: a new life for an old method]

Abstract

In this paper, I present the main approaches used in gene mapping of complex human diseases by linkage analysis based on molecular markers. The first section describes the traditional LOD-score analysis and gives a review of different improvements and refinements of this approach, which has been proposed in order to take into account complicating factors such as incomplete penetrance, genetic heterogeneity and unknown mode of inheritance. The second section describes the three main approaches of non-parametric linkage analysis, for which there is no need to specify a genetic model (mode of inheritance, allele frequencies, ..). A comparison between these three methods, which may seem to be more appropriate for complex diseases, is given based on the most recent published studies. In the third section and discussion, I compare the LOD-score and non-parametric methods by stressing the advantages and drawbacks of each approach as found in recent publications. I deduce that, in spite of its apparent and assumed inappropriateness to the analysis of complex diseases, the LOD-score method is still very useful and could provide, in some circumstances, more power and precision than model-free methods.