(146i) Multiscale Modeling for DNA Simulations

Numerous experimental and theoretical studies have offered clues regarding the molecular recognition capabilities of DNA, but our understanding of many aspects of the self-assembly process is still far from complete.  The goal of our DNA model is to help improve the fundamental understanding of the physical principles that underlie the DNA self-assembly process and to provide a computational tool based on molecular-level simulations that will enhance our ability to study the formation of DNA based nanostructures.  We used a multiscale modeling approach to develop an implicit-solvent intermediate-resolution model for DNA molecules designed for use with discontinuous molecular dynamics simulations and applied to study DNA structural properties and the hybridization process in solution.  The geometric and energetic parameters for this model were obtained by collecting data from atomistic simulations of single stranded DNA molecules with explicit solvent and counterions.  In this model, the sugar, phosphate and base in each nucleotide are each represented by a single coarse-grained site.  Using this new model we are able to simulate the spontaneous hybridization of 2 DNA strands composed of 24 base pairs in minutes. Simulation results show that the model accurately reproduces the hybridization of complementary DNA strands in mixtures containing both complementary and non-complementary DNA strands.  Current work is focused on using the model to predict the formation of Holliday Junctions.