The purpose of this project is to develop methodology for analyzing molecular population genetics data. Work has focused on the use of nonparametric methods for localizing susceptibility loci for complex diseases in humans. A strategy for mapping disease loci is to test for association with case-control samples, and follow up with a family-based test of association to confirm positive results. One way to improve the power of the first test is to increase sample size by combining the case-control and family data. To deal with the correlation between the two tests that this strategy introduces, we have developed a Monte Carlo procedure that always gives a valid test. For late-onset diseases it is common that parental genetic data are not available. Three recent family -based tests of association and linkage utilize an unaffected sibling as a surrogate for untyped parents. We have extended one of these tests and have compared the properties of the four tests in the context of a complex disease for both biallelic and multiallelic markers, as well as for sibships of different sizes. We have also examined the consequences of having some parental data in the sample. Two family-based tests of association and linkage to quantitative traits were developed that do not use parental data. One procedure assumes a mixed effects model in which the sibship is the random factor, the genotype is the fixed factor and the continuous phenotype is the dependent variable. Covariates can be easily accommodated and the procedure can be implemented using available statistical software. The second is a permutation-based procedure. We conducted a simulation study to illustrate the relative power of each test for a variety of quantitative genetic models. Optimal marker selection for the transmission/disequilibrium test (TDT) when applied to admixed populations was investigated. We found that collapsing a microsatellite marker to two alleles can increase the power of the TDT, and a method was developed for finding the optimal collapsing the uses estimates of alleles frequencies in the admixed population.
