Admixture in three Hispanic populations: Ancestry proportions, population structure, and gene mapping (Mexico, Colorado, New York)

Admixed populations constitute helpful natural experiments to understand evolutionary and epidemiological processes. Additionally, since admixture will create linkage disequilibrium (LD) between a disease risk allele and a marker locus, which will decay as a function of their distance and of the time since the beginning of admixture, recently admixed populations can be used to identify and map genes involved in complex trait inheritance as long as marker and trait differ substantially in frequency between parental populations. Admixture Mapping (AM) is a method that was developed to take advantage of this extended LD in admixed populations for the identification of complex trait genes. In Latin America, populations that were geographically isolated were brought together during the conquest and colonization of the American continent where they gave rise to a number of mixed populations with varying degrees of European, African, and Native American contributions. In the United States these populations are referred to as Hispanic or Latino, based on a definition created by the US Census Bureau. Despite the unifying label, Hispanic populations exhibit many differences related to culture, demography, socioeconomic status and genetic background. For this reason, this project advocates the study of particular Hispanic populations from a perspective that incorporates geographic and historical contexts. Three different Hispanic populations were studied: the Hispanic population of the San Luis Valley in southern Colorado which consists basically of Spanish Americans and Mexican Americans, a Puerto Rican population of women living in New York City, and the population of the city of Tlapa in the state of Guerrero, Mexico. Admixture proportions, extent of linkage disequilibrium, presence of population structure, and sex-biased gene flow were estimated in each case. To accomplish this goal a panel of autosomal, Y-linked and mitochondrial ancestry informative markers (AIMS) with large allele frequency difference between parental populations was assembled. In addition, the association of several complex phenotypes such as skin pigmentation, body mass index, and bone mineral density, with ancestry levels was assessed. Marker genotypes were also tested for association with these phenotypes. Skin pigmentation was significantly associated with ancestry in the San Luis Valley and the Puerto Rican populations, however, the markers that appeared associated with the trait were different in the two groups, suggesting the involvement of multiple candidate genes. Bone mineral density, but not body mass index, was correlated with ancestry in the Puerto Rican population. As expected, significant heterogeneity was found among these Hispanic populations as a result of differing histories of admixture and evolution following the admixture event. This variability will definitely influence the possibility of using these populations for AM of genes that underlie ethnic differences in complex disease/trait prevalence.