PAM: Particle Automata in Modeling of Multiscale Biological Systems

Serious problems with bridging multiple scales in the scope of a single numerical model make computer simulations too demanding computationally and highly unreliable. We present a new concept of modeling framework that integrates the particle method with graph dynamical systems, called the particle automata model (PAM). We assume that the mechanical response of a macroscopic system on internal or external stimuli can be simulated by the spatiotemporal dynamics of a graph of interacting particles representing fine-grained components of biological tissue, such as cells, cell clusters, or microtissue fragments. Meanwhile, the dynamics of microscopic processes can be represented by evolution of internal particle states represented by vectors of finite-state automata. To demonstrate the broad scope of application of PAM, we present three models of very different biological phenomena: blood clotting, tumor proliferation, and fungal wheat infection. We conclude that the generic and flexible modeling framework provided by PAM may contribute to more intuitive and faster development of computational models of complex multiscale biological processes.