Coarse-Grained DNA Modeling

The objective of this research is to develop a simplified and accurate computational model for DNA behavior that can be used to provide molecular level understanding of experimentally observed phenomena. The ultimate goal is to apply this predictive DNA model for improving design principles in nanotechnology applications such as sequence-dependent separation of carbon nanotubes from a mixture and DNA-mediated particle assembly to create novel hierarchical structured materials. In principal, fully detailed all-atom models that account for every atom in the system including solvent atoms can provide the most accurate representation. These models, however, have computational limitations both in system size and timescale of interesting processes that can be reached. In order to observe the phenomena of interest, a condensed model is required. We propose a coarse-grained model in which each nucleotide is represented by only two spherical beads – for backbone and base atoms. In this model, the effective interactions between base beads are derived from all-atom simulation data for both base-base stacking and base-base hydrogen bonding. (1) The model, which is not built around a particular reference state unlike other coarse-grained DNA models, is able to describe the structural and thermodynamic properties of both single and double strand DNA including hairpin and duplex formation. (2) Shankar, Jagota, and Mittal. \"DNA Base Dimers Are Stabilized by Hydrogen-Bonding Interactions Including Non-Watson–Crick Pairing Near Graphite Surfaces.\" The Journal of Physical Chemistry B 116.40 (2012): 12088-12094. Boyer, Ding, and Mittal. "A Novel Coarse-Grained Model for dsDNA and ssDNA." (to be submitted)