Digital quantification of human eye color highlights genetic association of three new loci

Abstract

Previous studies have successfully identified genetic variants in several genes associated with human iris (eye) color; however, they all used simplified categorical trait information. Here, we quantified continuous eye color variation into hue and saturation values using high-resolution digital full-eye photographs and conducted a genome-wide association study on 5,951 Dutch Europeans from the Rotterdam Study. Three new regions, 1q42.3, 17q25.3, and 21q22.13, were highlighted meeting the criterion for genome-wide statistically significant association. The latter two loci were replicated in 2,261 individuals from the UK and in 1,282 from Australia. The LYST gene at 1q42.3 and the DSCR9 gene at 21q22.13 serve as promising functional candidates. A model for predicting quantitative eye colors explained over 50% of trait variance in the Rotterdam Study. Over all our data exemplify that fine phenotyping is a useful strategy for finding genes involved in human complex traits.

Figure 1. The Hue-Saturation (H-S) eye color space in the Rotterdam Study (RS123).

(A) The Hue-Saturation (H-S) eye color space; (B) Example eye photos at their respective position in the H-S space; (C) 3 color categories defined by an ophthalmologist during eye examination and highlighted in the H-S space; (D) 5 categories defined by two researchers from digital full eye size photographs used for digital quantitative extraction of eye colors; (E) 4 quartiles of the 1st principle component C_HS1; and (F) 4 quartiles of the 2nd principle component C_HS2; all depicted on the H-S color space.

Figure 2. Observed and expected P values for eye color in the Rotterdam Study (RS123).

Observed −log₁₀ P values in a GWA of C_HS1 are ranked on the y-axis and plotted against the expected distribution under the null on the x-axis. All P values smaller than 10⁻¹⁰ were truncated at 10 at the log scale. The red dots are the P values excluding the effects of sex, age, and population stratification. Blue dots are the P values excluding the effects of 7 genes previously known to be involved in eye color. Green dots are the P values after additionally excluding the effects of 3 newly identified loci, with no more SNPs showing significant association at the genome-wide level.

Figure 3. GWA Manhattan plot for quantitative and categorical eye color in the Rotterdam Study (RS123).

The −log10 P values for association with 6 eye color traits (hue, saturation, C_HS1, C_HS2, 3-category classification, 5-category classification) are plotted for each genotyped SNP according to its chromosomal position (NCBI build 36). The distance between each tick on the x-axis represents 50 Mb. The P values smaller than 10⁻¹² are truncated at the level of 12 at the log scale and those greater than 0.01 are not shown. (A) P values are adjusted for age, sex and population stratification, (B) additionally adjusted for the effect of rs12913832 in HERC2, the most significantly associated eye color SNP known before, and (C) additionally adjusted for the effect of all 7 previously known eye color associated genes. Previously known eye color genes with genome-wide significant eye color association in the present study are noted using blue text above the figure and genes in the three newly identified loci in the 3rd scan are noted in red.

Figure 4. Chromosome 1q42.3 associated with quantitative eye color in the Rotterdam Study (RS123).

Regional association plots for 300 Kb surrounding the three newly identified eye color locus on chromosomal 1q42.3. Statistical significance of associated SNPs at each locus are shown on the −log(P) scale as a function of chromosomal position. P values were derived for 6 eye color traits (see figure legend). Genes in the region and LD patterns according to HapMap version 21a CEU samples are aligned bellow. Chromosome 1 233.85–234.25 Mb region includes the LYST gene, where SNPs rs3768056 and rs9782955 showed genome-wide significant association with saturation only.

Figure 5. Chromosome 17q25.3 associated with quantitative eye color in the Rotterdam Study (RS123).

The chromosome 17 77.05–77.35 Mb region includes multiple small genes, SNPs rs7219915, rs9894429, and rs12452184 showed genome-wide significant association with multiple traits.

Figure 6. Chromosome 21q22.13 locus associated with quantitative eye color in the Rotterdam Study (RS123).

The chromosome 21 37.30–37.65 Mb region includes DSCR6, PIGP, TTC3, DSCR9, and DSCR3 genes, SNPs rs1003719, rs2252893, rs2835621, rs2835630, and rs7277820 showed genome-wide significant association with C_HS1.