Abstract
Genotype-to-phenotype association studies typically use macroscopic physiological measurements or molecular readouts as quantitative traits. These approaches have been successful for the identification of disease risk loci, and variants that affect gene transcription and regulation. However, there are comparatively few suitable quantitative traits available between cell and tissue length scales, a limitation that hinders our ability to identify variants affecting phenotype at many clinically informative levels. We overcome these limitations by showing that unbiased quantitative image features, automatically extracted from histopathological imaging data, can be used successfully for image Quantitative Trait Loci (iQTL) mapping and disease variant discovery. Using thyroid pathology images, clinical metadata, and genomics data from the Genotype and Tissue Expression project (GTEx), we establish and validate a quantitative imaging biomarker for thyroid autoimmune disease. A total of 80,791 candidate variants in 769 coding genes highly associated with lymphocyte invasion in our analysis were selected for iQTL profiling, and tested for genotype-phenotype associations with our quantitative imaging biomarker. Significant associations were found with variants in Histone Deacetylase 9 (HDAC9), a gene proposed to be an epigenetic switch in T-cell mediated autoimmunity, but not previously associated with thyroid autoimmune disease. We validated our results using an independent dataset of 1,213 hypothyroidism cases and 3,789 controls from the Electronic Medical Records and Genomics network (eMERGE).
One Sentence Summary We use a histopathological image QTL analysis to identify genomic variants associated with thyroid autoimmune disease.