Artificial neural network approaches for modelling complex biological network – Mammalian cell cycle : A thesis submitted in partial fulfilment of the requirements for the Degree of Master of Applied Science at Lincoln University

Abstract

An important process in the growth of any biological organism is its ability to proliferate, a tightly controlled process in which a cell divides into two genetically identical daughter cells. This happens within a dynamic environment, where a cell responds to various internal and external signals through a well-ordered sequence of events called the cell cycle. Underlying these events is a complex and elegantly orchestrated web of interactions that function as an integrated system with various sub-systems that specialise in various tasks. Two such important tasks include cell cycle initiation in response to proliferative signals and the interaction of numerous elements for the completion of the cell cycle. This results in a highly complex system. Any malfunctioning during cell cycle division can cause diseases like Cancer. For gaining insights into biological reactions and their effects, cellular modelling approaches have contributed immensely. A few gaps are recognised in the field after reviewing the literature on mammalian cell cycle modelling. Most models are based on mathematical formulation representing the dynamic behaviour of the cell cycle which includes varied equations ranging from a few to tens of equations. They produce accurate systems dynamics, but the models are complex to solve and require the knowledge of many parameters. On the other hand, Discrete Models are simpler and use a qualitative approach but have numerous limitations to represent the continuous dynamics of the Mammalian Cell cycle. Therefore, there is a need for a modelling approach that is simplified but comprehensively represents the system. Mainly, the representation of a complex system in a robust way is a crucial demand. Our research mainly aims to introduce Artificial Neural Network approaches that mimic the mammalian cell cycle in an intuitive way. The goal is to explore the updated biological knowledge and develop ANN-based mathematical models to check their capabilities for mimicking cell signalling mechanisms.