Detecting Linkage between a Trait and a Marker in a Random Mating Population without Pedigree Record

Modern linkage-based approaches employing extended pedigrees are becoming powerful tools for localizing complex quantitative trait loci. For these linkage mapping methods, it is necessary to reconstruct extended pedigrees which include living individuals, using extensive pedigree records. Unfortunately, such records are not always easy to obtain and application of the linkage-based approaches has been restricted. Within a finite population under random mating, latent inbreeding rather than non-random inbreeding by consanguineous marriages is expected to occur and is attributable to coalescence in a finite population. Interestingly, it has been revealed that significant random inbreeding exists even in general human populations. Random inbreeding should be used to detect the hidden coancestry between individuals for a particular chromosomal position and it could also have application in linkage mapping methods. Here we present a novel method, named finite population based linkage mapping (FPL) method, to detect linkage between a quantitative trait and a marker via random inbreeding in a finite population without pedigree records. We show how to estimate coancestry for a chromosomal position between individuals by using multipoint Bayesian estimation. Subsequently, we describe the FPL method for detecting linkage via interval mapping method using a nonparametric test. We show that the FPL method does work via simulated data. For a random sample from a finite population, the FPL method is more powerful than a standard pedigree-based linkage mapping method with using genotypes of all parents of the sample. In addition, the FPL method was demonstrated by actual microsatellite genotype data of 750 Japanese individuals that are unrelated according to pedigree records to map a known Psoriasis susceptible locus. For samples without pedigree records, it was suggested that the FPL method require limited number of individuals, therefore would be better than other methods using thousands of individuals.