On the use of process algebra techniques in computational modelling of cancer initiation and development

Cancer research has been revolutionised by recent technological advances that allow scientists to produce extensive collections of experimental data, especially on the molecular and cellular level. Formal modelling is a necessary tool for integrating massive amounts of diverse measurement data into a coherent picture of disease development. Models can be used to test hypotheses about the role of cellular components in system function and in creating disease, and to make predictions which can then be tested experimentally. This thesis evaluates process algebra techniques as description formalisms for a collection of cancer-related models. Process algebras view biology as a dynamic interactive communication network, in which an individual agent is performing a computation corresponding to the reaction. Agents typically represent entities such as molecules or cells. The stochastic extensions of process algebras allow the modeller to assign probability (or rate) to every reaction. The analysis of the resulting models is usually based on stochastic simulation. Alternatively, formal verification tools can be used to calculate exact quantitative properties of the underlying stochastic process. We have explored the applicability of the process algebra formalism by analysing the dynamics of two cancer-related signalling pathways: Wnt/Wingless and FGF (Fibroblast Growth Factor). In addition to process algebra models, we have also derived continuous differential equation models for comparison. Systematic analysis of parameter spaces has revealed which variables have the most influence on temporal and steady state properties of the system. By integrating feedback mechanisms, amplification factors, and different time scales we have demonstrated a resulting emergence of several unexpected properties of system dynamics. We were later able to confirm these by in vitro experiments for both pathways. To examine the function of the specific signalling architecture in the cellular decision making process, we have constructed a model that couples Wnt signalling to the decision process within the cell and cell microenvironment. The model reveals signalling characteristics that ensure accuracy and robustness of Wnt-mediated determination of proliferative cell fate and lead to tissue architecture which is resistant to mutations. The main contribution of this thesis is, therefore, to systems biology; we have produced reusable and validated quantified models and demonstrated their value in designing, testing, and refining hypotheses about cancer.EThOS - Electronic Theses Online ServiceGBUnited Kingdo