iGWAS: Image-based genome-wide association of self-supervised deep phenotyping of retina fundus images

• Published in: PLoS genetics 2024-05, Vol.20 (5), p.e1011273-e1011273
• Main Authors: Xie, Ziqian, Zhang, Tao, Kim, Sangbae, Lu, Jiaxiong, Zhang, Wanheng, Lin, Cheng-Hui, Wu, Man-Ru, Davis, Alexander, Channa, Roomasa, Giancardo, Luca, Chen, Han, Wang, Sui, Chen, Rui, Zhi, Degui
• Format: Article
• Language: English
• Subjects: Biobanks; Biology and Life Sciences; Color vision; Computer and Information Sciences; Datasets; Deep learning; Design; Diagnostic imaging; Gene loci; Genetic factors; Genetic research; Genome-wide association studies; Genomes; Genomics; Genotype & phenotype; Image processing; Machine learning; Medical examination; Medical imaging; Medical imaging equipment; Medicine and Health Sciences; Methods; Neural networks; Phenotype; Phenotypes; Phenotyping; Quality control; Research and Analysis Methods; Retina
• ISSN: 1553-7404; 1553-7390; 1553-7404
• DOI: 10.1371/journal.pgen.1011273

Existing imaging genetics studies have been mostly limited in scope by using imaging-derived phenotypes defined by human experts. Here, leveraging new breakthroughs in self-supervised deep representation learning, we propose a new approach, image-based genome-wide association study (iGWAS), for identifying genetic factors associated with phenotypes discovered from medical images using contrastive learning. Using retinal fundus photos, our model extracts a 128-dimensional vector representing features of the retina as phenotypes. After training the model on 40,000 images from the EyePACS dataset, we generated phenotypes from 130,329 images of 65,629 British White participants in the UK Biobank. We conducted GWAS on these phenotypes and identified 14 loci with genome-wide significance (p