| Description |
Genetic constraint describes areas of the genome that do not tolerate change. Discovering constraint is necessary for furthering the knowledge of the functional components of the genome, and how they relate to human disease. Additionally, genetic changes, or variants, that overlap these regions of constraint may affect transcription, protein function, or active binding sites. Thus far, there have only been gene-wide models of constraint, and other tools for assessing genetic variation are wide-ranging and often in contention with one another. Evaluations of these tools have been inconstant. This dissertation attempts to improve the resolution of genetic constraint and create a consistent framework for evaluating tools that claim to determine the damaging nature, or pathogenicity, of genetic variants. In Chapter 1, the background necessary to understand the origins and current state of constraint and genetic variant evaluations is detailed. The paper presented in Chapter 2 details the first map of regional constraint in the human genome. This work is an improvement upon all previous models of constraint, and provides potential for new disease gene discovery in the context of autosomal and X-linked dominant diseases. In Chapter 3, a framework has been built for properly and consistently evaluating and conglomerating variant pathogenicity metrics. This creation acts as both a system for evaluating these metrics as well as a repository for pathogenicity scores and variant sets on which to evaluate them. Lastly, the impact of the work in Chapters 2 and 3, the new research that has already been performed that builds on this work, and the iv discussion of further improvements that can be made is contained within Chapter 4. Altogether, this dissertation expands our knowledge of the most crucial parts of the genome and our understanding of genetic variation's role in human disease. |
