Structural modeling of the 3D genome using machine learning

Abstract

This dissertation, submitted as a partial requirement for completion of the Doctorate of Philosophy, outlines the research performed by Max Highsmith in the BDM Lab. This work includes a functional expansion of a three-dimensional genome conformation database, the development of a novel, deep-learning based strategy for the enhancement of Hi-C data, The development of deep learning approach for domain identification using epigenetic features, and the development of a novel computational tool for 4D modeling of chromosome dynamics.