Genetic architecture of common and rare diseases

Abstract

We studied the genetic architecture of common and rare diseases in order to gain more insight into the underlying etiology. In contrast to epidemiological studies, genetic variants are generally not subject to confounding, allowing us to test causal genotype-phenotype relationships. However, it has become increasingly clear that study sample sizes need to be large in order to achieve sufficient power to identify robust associations with disease. This thesis is divided into three parts. In part one, we studied the genetic associations for three immune-related diseases. We searched for genetic variants that cause birdshot choreoretinopathy, a rare eye disease with a strong HLA component, and showed that the HLA-A*29:02 allele is the principal effect with more than 95% of patients carrying this allele. In addition, we identified an association near ERAP2, which is a gene involved in antigen processing and presentation. Second, we compared the genetic basis for multiple sclerosis and amyotrophic lateral sclerosis. Combining data collected in almost 20,000 samples in total, we failed to observe evidence for a shared genetic background between these two devastating diseases. In part two, we studied the genetic basis of arterial calcification as a risk factor for coronary artery disease and myocardial infarction. We conducted a genome-wide association study on coronary artery calcification and aorta calcification, and demonstrated that three known loci associated with myocardial infarction are also involved in arterial calcification. In addition, many additional SNPs below genome-wide significance are also associated with coronary artery calcification, highlighting an important causal role for arterial calcification in the etiology of myocardial infarction. Part three includes four studies focusing on quantitative phenotypes derived from the electrocardiogram. We tested common SNPs for association with PR interval and with four correlated QRS measurements in meta-analyses containing up to 100,000 individuals, and identified in total more than 100 genomic loci. In a smaller sample of 30,000 individuals we illustrated that imputation of untyped SNPs using larger and more accurate panels allows us to identify even more loci. In the last chapter, we tested low-frequency coding variants for association with PR, RR, QT, and QRS intervals. Although the sample size of our cohorts was too small to identify significant association signals, this study taught us important lessons for association testing of rare DNA sequence variants. To conclude, we applied different analytical strategies to investigate the genetic architecture of common and rare diseases, testing both common and rare variation. The progress of the last years in the field of genetics has brought us a tremendous amount of information on disease biology. We predict that GWAS and sequencing-based studies will continue to identify more and more loci that can serve as valuable starting points for eventual prediction, prevention, and treatment of human disease.