Exploring functional elements and genomic variation in the noncoding genome

Abstract

Gene expression regulation is a delicate process that depends on multiple aspects including genome structure and transcription factor binding to DNA elements. The majority of our genome consists of noncoding DNA, which was shown to be crucial in providing the correct context for genome function. Although DNA mutations in protein-coding genes have mostly predictable effects, it is still largely unclear how noncoding DNA variants affect genome function and if they contribute to common disease. Effects of noncoding mutations are more diverse, which makes their (combined) impact difficult to assess. For example, noncoding mutations might affect transcription factor binding to regulatory elements such as enhancers or promoters, target noncoding genes or influence the 3D organization of the chromatin. DNA, RNA and protein sequencing techniques now allow us to assess multiple levels of genome function. This includes all variation present in genomes (both at the structural and the single nucleotide level), quantitative and qualitative RNA and protein measurements, high-resolution assessment of chromatin states and genome-wide transcription factor binding profiles. However, correct interpretation and integration of these different data modalities, a so-called "multi-OMICs approach", remains challenging. To determine the multi-level effects of genomic variation and define disease-linked variants, interpretation and integration challenges need to be overcome. In the chapters presented in this thesis, we focus on the interpretation and integration of multiple types of data to achieve better insight in the role of the noncoding genome in genome function. To that extent, we explore the effects of different types of noncoding genomic variation on phenotypic diversity and disease. We find that noncoding variants have widespread effects that cannot be predicted by DNA information alone. We show that the integration of multiple sequencing techniques can provide detailed insight into the effects of variants on chromatin state and gene transcription. If used correctly, a multi-OMICs approach is not only valuable for improving insight in fundamental biological processes, but also for better patient diagnosis. In a clinical setting, multi-OMICs information facilitates the identification of disease-linked variants and helps to elucidate the molecular mechanisms that couple variants to disease.