Document Type


Publication Date



OBJECTIVES: Our objective is to assess the informativeness of dental morphology in estimating biogeographic ancestry in African Americans.

MATERIALS AND METHODS: The data are 62 dental morphological traits scored as nondichotomized and dichotomized in 797 individuals, 992,601 SNPs from 271 individuals, and 645 STRs from 177 individuals. Each dataset consists of Africans, Europeans, and African Americans. For each dataset, we summed Fisher Information (FI), then used STRUCTURE to estimate ancestry.

RESULTS: Total FI was highest for SNPs, followed by STRs, nondichotomized dental traits, and dichotomized dental traits. For both genetic datasets, Africans and Europeans fell into two distinctive clusters with low 90% credible regions for individual ancestry estimates. In African Americans, membership in the African cluster was 76.4% and 80.4% for SNPs and STRs, respectively. For the dental data, all Africans and Europeans had appreciable membership in both clusters and comparatively high 90% credible regions for individual ancestry estimates. Nonetheless, African Americans had consistently higher membership in the same cluster in which Africans had high membership. African American membership in this cluster was significantly higher for the nondichotomized form than for the dichotomized.

DISCUSSION AND CONCLUSIONS: FI potentially provides a useful gauge of the effectiveness of dental and genetic data for ancestry estimation. The comparatively high FI of nondichotomized dental traits suggests data in this form may be better suited for studies of admixture than dichotomized data. Because of high error in individual ancestry estimates, dental morphological data may be unable to distinguish differences in ancestry among individuals within admixed populations.