Show simple item record

dc.contributor.authorAlsharaiah, Mohammad
dc.date.accessioned2018-11-27T21:45:16Z
dc.date.available2018-11-27T21:45:16Z
dc.date.issued2018
dc.identifier.urihttps://hdl.handle.net/10182/10371
dc.description.abstractBiological systems such as cell cycle are complex systems consisting of an enormous number of elements. These elements interact in ways that produce nonlinear and complex systems behaviour such as oscillations. A number of modelling approaches have been used to explain these kinds of systems; they are classified into four classes (continuous, discrete, stochastic and hybrid). Ordinary Differential Equations (ODEs) are used to mimic the continuous dynamic behaviour of system components, while discrete models can simulate the biological elements as straightforward binary variables providing a qualitative view of system behaviour. Stochastic models are used to model the effect of noise in biological systems. Combined methods together introduce hybrid models to cover the limitations of individual models and take advantage of their strengths. This study introduces a series of advanced models with increasing resolution from discrete to continuous in a systematic way to model the mammalian cell cycle system. Each model includes the essential controllers of mammalian cell cycle. Specifically, they reveal cell regulators that control cell cycle transitions from one phase to another in cell division. In the first model, this work introduces a new biological network based on a qualitative approach-Boolean network. A new Boolean model with 13 proteins can capture the essential aspects of cell cycle phases and it can simplify cell cycle control system. The developed new, simple and intuitive Boolean model can mimic the fluctuation of Cyclins during cell cycle. Furthermore, it can show cell cycle phases based on the changes in Cyclins since each Cyclin leads the transition of cell cycle phases. Also, some important proteins that are missing in most of the current models, such as SCF ubiquitin ligase and c-Myc, have been added to our model. This study provides a deep understanding of the mechanism of mammalian cell cycle regulation, especially when growth factors are present, and it can reveal the cell cycle process in terms of flow cytometry of Cyclin proteins. In addition to the aforementioned advancements related to the developed Boolean model, the model has captured the periodic sequence of activity of cell cycle regulatory proteins over repeated cell cycles. The existing discreet models failed to represent the periodic behaviour of cell cycle phase transitions and the correct activities for system species over repeated cell cycles. In the second model, another computing model based on degree of truth rather than the usual Boolean model with true or false (1 or 0) values is implemented. This second developed model is a fuzzy logic model. It involves the use of artificial intelligence to model a fuzzy cell cycle control system. Fuzzy logic model is a rule-based model that depends on the practical knowledge and heuristic design of complex systems. Specifically, this work utilizes fuzzy inference system features in the development process. Indeed, it is expected to provide approximate continuous dynamics for the discrete events using limited available data. When applied to the cell cycle system, the model reveals the intermediate states of concentrations of proteins during each cell cycle event. In addition, applying fuzzy logic provides more accurate representation of the cell cycle system than previous Boolean approaches. It can follow/explain the flow cytometry measurement levels of protein concentrations such as Cyclins. Also, the fuzzy model is expected to simplify and realistically represent the system and address the existing shortcomings of both discrete and continuous representations, specifically, discrete models’ restriction to 0 and 1 values for the states of proteins and the scarcity of available kinetic parameters in continuous models such as ODEs. Furthermore, for more accurate results and to automate the development of the fuzzy controller, the core of the fuzzy inference system has been successfully optimized using artificial intelligent approach Particle Swarm Optimization (PSO).en
dc.language.isoenen
dc.publisherLincoln Universityen
dc.rights.urihttps://researcharchive.lincoln.ac.nz/page/rights
dc.subjectmammalian cellen
dc.subjectcell signalling networksen
dc.subjectsynchronous and asynchronous boolean networksen
dc.subjectattractorsen
dc.subjectfuzzy systemsen
dc.subjectparticle swarm optimization (PSO)en
dc.subjectcell cycleen
dc.titleQualitative and semi quantitative systems approaches to complex systems modelling – Intuitive and flexible models of biological systemsen
dc.typeThesisen
thesis.degree.grantorLincoln Universityen
thesis.degree.levelDoctoralen
thesis.degree.nameDoctor of Philosophyen
lu.thesis.supervisorSamarasinghe, Sandhya
lu.contributor.unitDepartment of Informatics and Enabling Technologiesen
dc.rights.accessRightsRestricted item - embargoed until 30 November 2019
dc.subject.anzsrc08 Information and Computing Sciencesen
dc.subject.anzsrc080301 Bioinformatics Softwareen
dc.subject.anzsrc080108 Neural, Evolutionary and Fuzzy Computationen
dc.subject.anzsrc060102 Bioinformaticsen


Files in this item

Default Thumbnail

This item appears in the following Collection(s)

Show simple item record