< Back to Article

Complex Patterns of Genomic Admixture within Southern Africa

Figure 2

Population substructure within the five southern African populations (n = 103) in relation to 10 globally relevant populations (n = 237) for 33,207 overlapping autosomal markers.

(A) Principal Components Analysis (PCA) clustering of the Yoruba, Ju/'hoan and Han Chinese/Europeans at the outer extremes of Coloured and Baster dispersal is suggestive of ‘African non-Khoesan’, ‘African Khoesan’ and ‘non-African’ substructure, while the amaXhosa and !Xun form tight clusters representing their Bantu and Khoesan ancestral roots, respectively. (B) STRUCTURE analysis (K = 3 to K = 9) further defines homogenous ancestral contributions defined in this study as ‘Khoesan’ (Ju/'hoan, green), ‘African proto-Bantu’ (Yoruba, red), Hadza (light blue), ‘Asian’ (Han Chinese, blue), European (yellow), and ‘Pygmy’ (light gray), the latter further defined (K = 7) as Mbuti (dark gray) and Biaka (maroon), Sandawe (light purple) and ‘Nilotic’ (Maasai, orange).

Figure 2