Table 1.
Characteristics of subjects.
Figure 1.
Multidimensional scaling plots for all samples (TOP) and only biobank samples (BOTTOM).
Plots of subjects in the 1st and 2nd dimensions (left column), 2nd and 3rd dimensions (middle column), and 3rd and 4th dimensions right column). Abbreviation for samples: African American biobank sample (ANY); European American biobank sample (ENY); Hispanic American biobank sample (HNY); African ancestry in Southwest USA (ASW); Mexican ancestry in Los Angeles, California (MEX); Yoruba from Nigeria (YOR); African American from Maywood, Illinois (AMW); Brazilians from Brazil (BRZ).
Figure 2.
Canonical correlation based on window-wide local PCs for biobank samples and selected external samples.
Each circle represents the squared coefficient of the largest canonical correlation between the first 10 local PCs of a local 20 Mb-window and the first 10 global PCs. Abbreviation for samples: African American biobank sample (ANY); European American biobank sample (ENY); Hispanic American biobank sample (HNY); African American from Maywood, Illinois (AMW); CEPH (Utah residents with ancestry from northern and western Europe (CEU); Brazilians from Brazil (BRZ).
Figure 3.
Population structure results for ancestral populations K = 2 to K = 6.
Each subject is represented by a thin vertical line colored in proportion to their estimated ancestry within each cluster. The colors represent the proportion of inferred ancestry from each of the ancestral populations within each specific K value. Abbreviation for samples: African American biobank sample (ANY); European American biobank sample (ENY); Hispanic American biobank sample (HNY); African ancestry in Southwest USA (ASW); Mexican ancestry in Los Angeles, California (MEX) ; Yoruba from Nigeria (YOR); African American from Maywood, Illinois (AMW); Brazilians from Brazil (BRZ).
Figure 4.
Population structure results for K = 2 ancestral populations sorted by ancestry proportions for African American biobank sample (ANY) and Hispanic American biobank sample (HNY) (Top) and pooled sample of both ANY and HNY (Bottom).
Figure 5.
a: Linkage disequilibrium structure (Top) and organization of haplotypes harboring published obesity variants (rsIDs indicated with blue boxes) in FTO gene in biobank sample of African Americans (left column), European Americans (middle column), and Hispanic Americans (right column). b: Linkage disequilibrium structure (Top) and organization of haplotypes harboring published obesity variants (rsIDs indicated with blue boxes) in MC4R gene in biobank sample of African Americans (left column), European Americans (middle column), and Hispanic Americans (right column).
Figure 6.
a: Linkage disequilibrium structure (Top) and organization of haplotypes harboring published pharmacogenomic variants (rsIDs indicated with blue boxes) in SLCO1B1 gene in biobank sample of African Americans (left column), European Americans (middle column), and Hispanic Americans (right column). b: Linkage disequilibrium structure (Top) and organization of haplotypes harboring published pharmacogenomic variants (rsIDs indicated with blue boxes) in CYP4F2 gene in biobank sample of African Americans (left column), European Americans (middle column), and Hispanic Americans (right column).
Figure 7.
BMI associations with published obesity variants in FTO (Top) and MC4R (Bottom) genes in the three biobank samples.