Strategies based on marker information for the study of human diseases

Publisher Website
Google Scholar
Abstract
The goals and the ways of using genetic marker information when studying human disease are very different according to whether the disease or sub‐entity of the disease is mendelian or if a ‘disease gene’ in the sense of a rare mutated allele does not exist but rather common genetic risk factors, each one normal if considered alone. In the former case, genetic markers are used in the aim of localizing the defective gene and a systematic screening of the genome seems to be an efficient strategy provided there is not too much ambiguity in the correspondence between phenotypes and genotypes. In the latter case, the goal is to find risk factors allowing us to predict better the risk for an individual and to define different risk groups resulting in greater power to show the potential role of other factors (genetic or environmental). In this situation, the use of the lod score method with random markers presents several disadvantages: first, the multiple testing problem is particularly crucial; second, false rejection of linkage may be induced by misspecification of the model describing the genetic basis of the disease; and last, the power of detecting linkage may be low. A strategy focusing on ‘candidate gene’ markers may be then more efficient.