METASIM: object-oriented modelling of cell regulation.

Enzymatic processes and substances are modelled as distinct objects, belonging to a limited number of classes. A set of class definitions in C++ is presented that constitutes an object-oriented programming platform. The latter supports 'biological' data types and functions and facilitates simulation of metabolic and regulatory pathways in living cells. To compute the time-evolution, Euler or Runge-Kutta methods are used, though the latter method compromises a strict object-oriented philosophy. As an example, histone gene expression during embryogenesis of Xenopus laevis is modelled. This object-oriented programming system forms a modelling 'language' which is readily understood by both biochemists and programmers. It allows biological problems to be programmed more easily and correctly and brings the program closer to the biological reality, hence making it more meaningful to bioscientists. Moreover, it can readily be extended to new models by class derivation.