Investigating the genetic architecture of eye colour in a Canadian cohort

Abstract

The main factors that determine eye colour are the amount of melanin concentrated in iris melanocytes, as well as the shape and distribution of melanosomes. Eye colour is highly variable in populations with European ancestry, in which eye colour categories cover a continuum of low to high quantities of melanin accumulated in the iris. A few polymorphisms in the HERC2/OCA2 locus in chromosome 15 have the largest effect on eye colour in these populations, although there is evidence of other variants in the locus and across the genome also influencing eye colour. To improve our understanding of the genetic loci determining eye colour, we performed a meta-analysis of genome-wide association studies in a Canadian cohort of European ancestry (N= 5,641) and investigated putative causal variants. Our fine-mapping results indicate that there are several candidate causal signals in the HERC2/OCA2 region, whereas other significant loci in the genome likely harbour a single causal signal (TYR, TYRP1, IRF4, SLC24A4). Furthermore, a short subset of the associated eye colour regions was colocalized with the gene expression or methylation profiles of cultured melanocytes (HERC2, OCA2), and transcriptome-wide association studies highlighted the expression of two genes associated with eye colour: SLC24A4 and OCA2. Finally, genetic correlations of eye and hair colour from the same cohort suggest high pleiotropy at the genome level, but locus-level evidence hints at several differences in the genetic architecture of both traits. Overall, we provide a better picture of how polymorphisms modulate eye colour variation, particularly in the HERC2/OCA2 locus, which may be a consequence of specific molecular processes in the iris melanocytes.

Decreased retinal vascular complexity is an early biomarker of MI supported by a shared genetic control

“…In addition, we completed a supplementary GWAS including an eye colour PRS based on the study by Lona-Durazo et al . 23 , which indicated no eye colour effect in our GWAS results (see Section “Methods”). The quantile-quantile plot of both GWASs indicated an adequate control of the genomic inflation in our analysis (λ GC = 1.065 and λ GC = 1.067 in the right and left eye, respectively, see Supplementary Fig.…”

“…This PRS derives from an eye colour GWA study that defines it quantitatively (i.e. 1 = blue or grey, 2 = green, 3 = hazel, and 4 = brown) completed by Lona-Durazo et al using the CanPath cohort, which includes ∼5000 participants with European ancestry 23 . We estimated this PRS for each participant by extracting those genetic variants with a P-value<5·10 −8 from the summary statistics and applying linear regression to the effects of these SNPs and the genotypes of our UKBB participants.…”

“…To investigate the shared genetic control between D f and associated traits, we estimated their genome-wide genetic correlation. For this purpose, we obtained the GWAS summary statistics of traits of interest to our study from GeneATLAS and the eye colour study 23 . These calculations were computed with LD Score, a toolbox that estimates genetic correlation using GWAS summary statistics considering possible inflation caused by SNPs in linkage disequilibrium (LD).…”