Bernd Mayer, Steen Rasmussen

Paper #: 96-08-059

Molecular self-assembly is frequently encountered in biochemical systems generating higher order structures with well-defined functionalities. However, the driving forces underlying these processes are not well understood. The "Lattice Molecular Automaton" (LMA) is a computational tool suitable for simulation of self-organization processes in large-scale, molecular systems. This paper introduces the basic computational concepts needed to formulate molecular dynamics and self-assembly in a discrete field, cellular automaton environment: Molecular objects are encoded as data structures on a hexagonal lattice. Propagating force particles, together with kinetic and potential energy terms, define simulation objects with a minimum complexity (number of physical variables together with interaction functions) with respect to specified molecular dynamics and force field properties. In this paper we focus on the mathematical and algorithmic formulation of a variety of intermolecular interactions through a decomposition of molecular type-specific force fields into discrete fields constructed by propagating force information particles. As an example, the simulation of polymer dynamics in an aqueous environment is shown. The straightforward implementability of the LMA concept on massively parallel architectures as well as possible applications in the field of Computational Nanotechnology are briefly discussed. Thermodynamical properties together with a variety of other physico-chemical properties of the LMA are discussed in detail in reference [1].