Peter Minary, Department of Structural Biology and Bio-X Institute Stanford University, USA

Towards Computational Structural Epigenetics at DNA/RNA Levels
Epigenetics, which is the study of changes in gene expression caused by mechanisms that preserve the underlying DNA sequence, has been an emerging field of research due to its promises to deliver novel treatments to genetic diseases. One of the major DNA level epigenetic changes, DNA methylation is expected to effect nucleosome positioning or chromatin structure, both influencing gene expression through controlling the accessibility of transcription on all DNA regions. Due to their dependence on sequence and/or structural libraries and coarse-grained modeling, the most successful computational methods for nucleosome occupancy lack the necessary physics based scaling and resolution to model delicate epigenetic changes such as DNA methylation. The complex repertoire of post-transcriptional regulation is dominated with changes arising at the RNA level. Such changes are but not limited to mRNA transport or RNA inference both involve large RNA structural assemblies rich in molecular junctions of various orders. Modeling the role and epigenetic function of these large RNA complexes is dependent upon fundamentally novel meso-scale all-atom resolution simulation technologies. In answer to this need, we present meso-scale all-atom simulation protocols to predict nucleosome occupancy along genomic sequences of kilobase length and sample RNA junctions of desired complexity. The robustness of these methods are demonstrated through predicting the in vitro experimental occupancy profile for a 20,000 base pair DNA sequence and the methylation effect on nucleosome occupancy. In addition, we present applications for sampling RNA junctions and probing their role in RNA based nanotechnology.

Bibliography