抄録
Dynamic computer simulations are essential for quantitative understanding of integrated mechanisms underlying various cell functions and can provide a powerful tool to create and test working hypotheses. We have been developing mathematical models of cell functions and biodynamics by integrating experimental data and implemented cardiac cell (Kyoto model from Matsuoka et al., 2003 and 2004, LRd model from Faber and Rudy, 2000, Noble et al., 1998), an epithelial cell (Strieter et al., 1990) and a pancreatic β cell model (Magnus and Keizer, 1998) using a common biological simulation tool called simBio, which is written in Java, uses XML and solves ordinary differential equations. In analogy to biological functional structures, a cell model in simBio is composed of independent functional modules called Reactors such as ion channels and sarcoplasmic reticulum, and dynamic variables called Nodes such as ion concentrations. Interactions between Reactors and Nodes are described by the graph theory and the resulting graph represents a blueprint of an intricate cellular system. Each Reactor can be composed or improved independently and can easily be reused for different models. We have designed simBio as a tool for a simple conversion of a biological system into machine code, for an easy model expansion, and for a reuse of models as a whole or in part as sub-models in different contexts. Thus, the development of a large variety of biological models will be enhanced. The simBio package is freely available from http://www.sim-bio.org/. [Jpn J Physiol 55 Suppl:S97 (2005)]
