Genomewide Association of Quantitative pigmentary traits in admixed US populations

As submitted by the proposer: Pigmentation of the skin, hair, and eyes shows remarkable levels of variation across the human species. The identification of the genetic loci that underlie variation in these externally visible characteristics (EVCs) provides exciting opportunities for advancement in the field of forensic science. An improved knowledge of the genetic markers responsible for this complex trait will make it possible to determine the probability of certain phenotypic characteristics (e.g. skin, hair, and eye color) using an unidentified sample of DNA. This ability to predict phenotype may ultimately be applied in directing police investigations by limiting suspect pools when other intelligence is not available, as well as in the identification of missing persons and human remains recovered from mass disaster sites. To date research in this field has been largely focused on identifying loci associated with normal variation in hair and iris phenotypes in populations of European ancestry. Without basic knowledge about the loci that underlie pigmentation variation in non-European populations, and particularly in admixed populations, it will be impossible to apply many of the advances of phenotype modeling in the field to the general US population. Further, many studies have relied on categorical, rather than quantitative descriptions of pigmentation phenotype. This results in a loss of statistical power to detect causal loci. The use of a quantitative approach to phenotype assessment will result in better power to detect loci of small effect. In this project we propose to take the first steps to remedy these deficiencies by focusing on phenotypic and genetic variation in two admixed populations. To that end, this proposal has four primary objectives 1) to quantitatively assess skin, hair, and iris phenotype in 500 individuals from each of two admixed US populations (African Americans and Hispanics), 2) to genotype these samples using population-specific Affymetrix Genome-Wide Axiom Microarrays specifically designed to capture common variation in admixed populations, 3) to impute untyped genotypes using the combined 1000 Genomes panel as a reference panel, and 4) to identify associations between directly genotyped and imputed markers and quantitatively measured pigmentation phenotypes using genome-wide association conditioning on individual ancestry and admixture mapping techniques. This work represents an important step forward in forensic science because a) it includes two under-studied populations that are a significant and growing part of the U.S. population, b) will provide quantitative assessments of the complete pigmentation phenotype, c) uses population-specific microarrays specifically designed to capture common variation in admixed populations, d) will use well-documented methods to control for variation in ancestry among admixed individuals, and e) will make it possible to identify loci responsible for pigmentation differences between major population groups (Europeans, West Africans and Native Americans). We anticipate that loci identified as part of this study may ultimately be of use in efforts to develop a panel of markers that could be used in forensic DNA phenotyping. While the information derived from such a panel should never be used to identify a specific suspect, inference of EVCs could be useful to guide the initial stages of a forensic investigation, or to help in the reconstruction of the appearance of deceased individuals for whom many physical characteristics cannot be determined.
ca/ncf