Mitochondrial variation in Taiwan Biobank reveals ancestry structure and trait associations

Figures

• 

  Figure 1. mtDNA variant statistics in Taiwan Biobank (A–C) Histograms display nDNA coverage (A), mtDNA coverage (B), and mitochondrial copy number (C) across 1,465 WGS samples. (D) Scatterplots display correlations of allele frequencies (AFs) between TWB and gnomAD: overall AFs (D) and East Asian (EAS) AFs (E). Gray dots represent shared variants between datasets, while blue and purple dots denote heteroplasmies and homoplasmies unique to TWB. The line of identity is depicted in each plot. Subplots below compare heteroplasmic and homoplasmic variants separately.

• 

  Figure 2. Mitochondrial haplogroup distribution and population structure in the Taiwan Biobank (A) A bar graph displays the prevalence of mtDNA haplogroups among TWB participants, arranged according to their phylogenetic relationships. Haplogroups deriving from macrohaplogroup M are colored in shades of blue, N in shades of green, and R in shades of purple. For clarity, each haplogroup’s color is consistent across all subfigures. (B) A simplified tree depicts mtDNA haplogroups found among TWB participants, with each haplogroup uniquely colored. Asian and European haplogroups are delineated by distinct colors as defined in MITOMAP. Haplogroups not present in the TWB are shown in gray. (C and D) PCA using nuclear variants from 1,492 participants, with individuals colored by self-reported parental ancestry (C) and haplogroup assignment (D). (E and F) PCA using mtDNA variants from 1,492 participants, with individuals colored by self-reported maternal ancestry (E) and haplogroup assignment (F).

• 

  Figure 3. Manhattan plots of mtDNA variant associations with complex traits (A) Results of associations between mtDNA variants and quantitative traits analyzed using generalized linear models (GLMs). Each dot represents an mtDNA variant ordered by genomic position, with colors denoting different phenotypes. (B) Results of associations between mtDNA variants and binary traits analyzed using GLMs, with colors denoting phenotypes. In both panels, the thin red horizontal line indicates the variant-adjusted significance threshold (p = 1.6 × 10−4). The thick orange line indicates the phenotype-adjusted threshold (p = 4.3 × 10−6 for quantitative traits and p = 3.4 × 10−6 for binary traits).

• 

  Figure 4. mtDNA variants associated with high myopia A Manhattan plot shows mtDNA variants tested for association with high myopia, ordered by their position on the mitochondrial genome. Variants are color-coded by genomic location: control region (green), intergenic region (orange), protein-coding region (purple), rRNA genes (pink), and tRNA genes (light green). The red dashed line indicates the variant-adjusted significance threshold (p = 1.6 × 10−4).

• 

  Figure 5. mtDNA variants associated with renal function and haplogroup lineages (A) A Manhattan plot illustrates associations of mtDNA variants with renal function biomarkers, including serum creatinine (Scr) levels and estimated glomerular filtration rate (eGFR). Variants are color-coded by genomic location as in Figure 4. The red dashed line indicates the variant-adjusted significance threshold (p = 1.6 × 10−4). (B) A heatmap displays the pairwise LD (r2) among the identified mtDNA variants, clustered into two main sets (orange and blue). (C) A phylogenetic tree shows the distribution of the mtDNA variants across haplogroups. Variants are color-coded according to the sets defined in (B), with haplogroup-defining variants not associated in the analysis shown in black.