Modelling biological systems stochastically is often necessary in order to
accurately represent  the fluctuations present in these systems. The
mathematical model is the master equation. Solving the master equation
numerically for multi-dimensional systems is however very computationally
demanding due to the 'curse of dimensionality'. The linear noise
approximation for the master equation can then reduce the computational
efforts.

The linear noise approximation for the master equation has in this project
been applied to enzymatic systems of two- and four dimensions. The
approximation lead to a separation of scales, giving differential
equations for the mean values of each involved molecular species and a
stochastic equation for the corresponding fluctuations. This makes it
possible to reduce the dimension of the problem. The equations were solved
numerically over time, displaying the dynamics of the system. For further
simplification, and to give a more exact approximation, a separation of
time scales was also performed.

The obtained results, steady-state values and probability distributions,
were in agreement with earlier studies of the same system in. The
computational time and complexity were largely reduced compared to earlier
numerical calculations, and the use of time-scale separation further
reduced the computational efforts and gave an increased accuracy of the
probability distribution. The results obtained here show the possibility
of stochastic in silico modeling of systems of high dimensionality, with
the linear noise approximation.