PROJECT SUMMARY A fundamental goal of human genetics is to identify and characterize deleterious mutations responsible for disease. Deleterious disease alleles are introduced by mutation, then increase or decrease in frequency through the effects of genetic drift and natural selection. Their ultimate fate depends critically on their effects on fitness in heterozygotes and homozygotes. Although there is clear evidence that humans carry a burden of deleterious mutations, the total number of such mutations and their typical fitness consequences remain unclear. Furthermore, it is unknown if mutations responsible for disease when homozygous also have subtle fitness effects when heterozygous. Minor fitness effects in heterozygotes may not present clinical symptoms, however can dramatically impact the frequencies of such alleles over evolutionary timescales. Finally, researchers have traditionally been limited to studying mutations that manifest their deleterious effects after birth and thus embryonic lethal alleles have remained largely unidentified. A substantial proportion of human genes appear to be under strong selective constraint, however have no known function, presumably because they are required for normal embryonic development. The goal of the proposed research is to address these unanswered questions in characterizing and mapping deleterious mutations in humans. Specifically, the proposed work combines bioinformatics, population genetics, and human genomics to estimate the burden of strongly deleterious recessive alleles, determine the fitness and dominance effects of mutations classified as recessive, and identify dominant and recessive embryonic lethal mutations. The first aim is to take advantage of autozygous regions in a founder population, the Hutterites, to estimate the deficiency of strongly deleterious genotypes and compare the result to a colony of vervet monkeys with a similar pedigree structure to understand differences among populations and species. The second aim is to develop a likelihood framework and use realistic simulations of human demographic history to test if the frequency distribution of deleterious alleles in genes annotated as completely recessive is consistent with the absence of any fitness effects in heterozygotes. The third aim is to sequence miscarriage samples and develop a bioinformatics pipeline to identify candidate mutations underlying embryonic lethality. Here, the bioinformatics pipeline will be made publically available for use by the clinical research community interested in understanding mutations responsible for spontaneous miscarriage. Together, the proposed research will address central questions about the population genetics of deleterious alleles and provide new insight into the forces that influence the persistence of human genetic disease. !