Functionality and sensing in Boolean networks

• Main Author: Luo, Jamie X.
• Published: University of Warwick 2012
• Subjects: 500; QH301 Biology
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.560342

The main theme of this thesis is investigating how a cell’s biological function relates to the topology of its Gene Regulatory Network (GRN). In this context, the limits a biological function places on evolution are examined and also whether genetic networks can evolve the capacity to sense internal mutations. GRNs are modelled using Threshold Boolean Networks (TBNs), abstracting away details so that certain computational approaches become viable. For instance in Chapter 3, all possible TBNs that attain a specified functional path (of the form {v(t)}T t=0) through the expression state space are exhaustively found from a possible 3N2 TBNs, where N is the number of genes (nodes) in the network. This allows for the detailed examination of the complete neutral evolutionary space of a given functional path. It is demonstrated that the major quantities of interest, such as the connectivity of this neutral space under point mutations, the mutational and noise robustness of the TBNs in this space and even the number of networks all depend strongly on the duration T of the paths. The neutral space is found to disintegrate rapidly into disconnected components as T is increased. The effect of more exotic functional forms is also investigated. Chapter 4 focuses on evolving networks which are sensitive to deletion mutations. It is found that increased sensitivity is readily evolvable in TBNs, with the networks evolving to be more topologically balanced (they possess a similar number of excitatory and inhibitory interactions). Networks are only found to achieve maximal sensitivity through attaining long limit cycles. The study of sensitivity is extended to static populations of TBNs in Chapter 5 and the question is asked about whether a population of cells can develop the capacity to sense the presence of a mutant among them. The multicellular framework is also used to investigate the effect of intercellular connectivity on the dynamics. It is found that the greater the intercellular connectivity the more uniform the expression patterns are between cells. Chapter 6 applies the general Ergodic Set (ES) [Ribeiro and Kauffman, 2007] concept to stem cell differentiation and cancer. An alternative hypothesis to that in [Serra et al., 2010] is proposed about how to model stem cell differentiation using ESs. Coupled with results from Chapter 4, I suggest that under this new hypothesis, pluripotent stem cells will correspond to more sensitive TBNs and that differentiated stem cells will correspond to more robust TBNs.