Prediction of the linked regions and exclusion probabilities—Requirement on family sizes in linkage analyses

Abstract

Transition of genotyping methods from the use of sparse microsatellite markers to the use of high-density SNPs in linkage analyses allows accurate determination of crossover points and allele sharing status among individuals in families. This facilitates extraction of full inheritance information and making full use of family members available. For well-defined inheritance models, linkage analyses based on single families of intermediate size is feasible because of the development in genotyping technology, and necessary since large families or multiple families with homogeneous genetic causes of unknown mutations become very rare. In this study, we developed an algorithm that can predict the fractions of genomic regions that can be excluded by linkage studies based on family structure and size. Software was also developed to allow evaluation of family sizes in linkage analyses based on crossover simulations and determination of allele sharing status among family members. The program can determine, given a certain family structure and size, inheritance model and penetrance: 1). the probability that a chromosome can be excluded from consideration and 2). size distribution of the regions not being excluded when the chromosome does not contain the mutation; 3). size distribution of the true region with the causal mutation that can be determined in that family based on a number of simulations. It provides a reliable prediction tool to help determine the sufficiency of a family for a linkage analysis, or portion of the genome that can be excluded from consideration when the size of a family is too small for a complete linkage analysis. The latter is important for identifying the mutation responsible for a given family for diseases of known causes but of enormous genetic heterogeneity, and also for mutation identification of a candidate gene approach, in narrowing the list of candidates need to be screened.