Case populations must match the respective disease model: Genotype diversity causes linkage disequilibrium mapping failure in monogenic disorders.

Abstract

Traditional linkage analysis in large families is the most promising approach for mapping disease genes of monogenic heritable disorders when the number of informative meioses is sufficient. With rare diseases, however, the low availability of informative pedigrees poses a significant limitation. As an adjunct to family linkage methods, association studies based on the investigation of individual haplotypes from a number of unrelated patients (i.e. linkage disequilibrium analysis) have recently been employed in mapping hereditary disease loci. However, such haplotype analysis is hampered by a number of effects that influence statistical evaluation, e.g. i) population history and size, ii) allele and haplotype frequencies in the respective population(s), iii) heterogeneous mutation and natural selection processes, and iv) small sample sizes of patient groups. The purpose of the present study was to determine the utility and limitations of haplotype-based genetic mapping in estimating the location of the NYX gene, which has recently been identified as the causative gene for a rare inherited retinal disorder known as the complete type of X-linked congenital stationary night blindness (CSNB1). For this purpose we recapitulated haplotypes and tested for linkage disequilibrium in 20 unrelated male CSNB1 patients from three European populations and 44 healthy individuals. All subjects were genotyped for 17 polymorphic microsatellite loci covering the Xp11.4 region with an average marker density of approximately 0.29 cM. We found that a precise model to describe mutations at loci that erroneously break up linkage is highly required, and that the case population must match the respective disease model.