(197af) Data-Driven Simulation of Chromatin Structure and the Epigenetic Code

The three-dimensional structure of DNA in the cell nucleus is increasingly recognized as an important factor in disease and development. Interestingly, chromatin organization has been correlated with a number of chemical modifications deposited "on top of" the genome (i.e. epi-), spawning the field of epigenetics. These epigenetic modifications do not alter any base pairs, but nevertheless constitute a one-dimensional code which has been hypothesized to template and regulate the three-dimensional folding of DNA. Such patterns of epigenetic marks and DNA structure are highly specific to cell type, organism development, and even disease state. Impressively, multiple experimental techniques have arisen to measure aspects of chromatin structure, such as pairwise contact frequencies. However, no experimental technique furnishes the complete three-dimensional structure. This is further complicated by the fact that chromatin exists in a statistical ensemble of structures, with high variability from cell to cell. We present results using techniques from statistical physics to reconstruct the most likely ensemble of chromatin states consistent with the experimental data. This workflow complements experiments in order to elucidate the structure of chromosomes at increased resolution. With these models, we investigate the sequence-to-structure relationship between epigenetic marks and chromosome structure and ask the question: do epigenetic marks alone carry sufficient information to template three-dimensional chromosome organization?