Abstract

The recent breakthroughs in biological experiments have enabled the researchers to measure the quantities of different chemicals that build biological units such as cells. This type of information can be used to build models that can explain and predict the behaviour of the system. Such models can later be used to design control mechanisms that can influence the behaviour of the system in a desired way. With the help of medicine and biology researchers, the designed control mechanisms can be translated into drugs that can control or cure major diseases.

Biological systems usually consist of complex networks of biological components that function through various reactions. In order to affect the behaviour of the system efficiently, the chemicals that have the highest influence on the system behaviour have to be found using a sensitivity analysis. Such chemicals, regarded as inputs, will be the targets for drug design (or other control actions).

Various modeling tools have been empolyed to capture the behaviour of biological systems. Perhaps the most widely used models are the ordinary differential equations (ODEs). In this thesis, an alternative model is propposed for the study of the chemical reactions based on stochastic Petri nets, one type of discrete event systems. It is shown that the proposed method can be used to find the changes of the chemical reactants. The advantage of the proposed method is that it is amenable to implementation on computing systems with parallel processors. This in turn reduces the (time) computational complexity (compared with ODE based simulations). The Petri net based simulations are also used to perform sensitivity analysis. The proposed method is illustrated using the Caspase Apoptosis network.