Innovations in Big Data and Omics Towards Phenome-Wide and Genome-Wide Association Studies for Retinal Disease

The human retina is a multilayered tissue with diverse embryological origins and is, therefore, a window for assessment of ocular, neurological, vascular, and other systemic conditions. Innovations in big data, particularly involving large population biobanks, have enabled improved data sets that facilitate a better understanding of connections between the retina and systemic health. Retinal fundus and optical coherence tomography imaging data provide rich resources for describing interpretable imaging endophenotypes and evaluating their utility as biomarkers for future ocular and systemic conditions. Furthermore, the addition of genomic data to retinal and ophthalmic information has enabled unbiased discovery of the biological mechanisms driving changes in the retina. Assessment of inherited common and rare variants influencing ocular phenotypes through genome-wide association studies, in silico analyses for gene prioritization, pathway enrichment analysis, and experimental validation may enable the identification of biological targets for therapeutic modulation of retinal microvascular indices, neuronal health, and other imaging biomarkers. Further development of polygenic risk scores for retinal phenotypes enables a personalized medicine approach to quantifying inherited disease risk for an individual. While interpretation of causality from observational associations is difficult, Mendelian randomization (MR) analyses utilizing genomics can help improve understanding of the causal relationship between different phenotypes. Given the routine, noninvasive nature of retinal imaging, the findings from retinal cross-phenotype and genome-wide analysis have the direct potential for application clinically including in diagnosis, monitoring and prevention, and in treatment of both ocular and systemic conditions.