Please use this identifier to cite or link to this item:
|Title:||Computational modelling of the biological and social factors of type 2 diabetes||Authors:||Dutta, Pritha||Keywords:||Science::Physics||Issue Date:||2021||Publisher:||Nanyang Technological University||Source:||Dutta, P. (2021). Computational modelling of the biological and social factors of type 2 diabetes. Doctoral thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/155229||Abstract:||Type 2 diabetes (T2D) is a metabolic disorder characterized by high levels of blood glucose resulting from defects in insulin secretion and insulin action. Of all the diagnosed cases of diabetes, T2D accounts for approximately 90% of the cases. The pathogenesis of T2D involves both biological factors as well as social factors. These factors have intricate interactions with each other, and to study the behaviour of such complex systems computational models are required. Computational models provide a framework to summarize existing knowledge, enable competing hypotheses to be compared qualitatively and quantitatively, and facilitate the interpretation of complex data. Moreover, computational models allow questions to be investigated that are difficult to approach experimentally. Theories can be tested in context, identifying the gaps in our understanding and potentially leading to new hypotheses. In this thesis, three computational models are proposed: a system dynamics model of the system of social norms regarding body weight perception and obesity prevalence, a Boolean network model of the insulin resistance pathway and pancreatic beta-cell apoptosis pathway, and an ordinary differential equation model of the system of signalling components involved in insulin gene expression and beta-cell identity. In addition, this thesis also includes two methods for inferring computational models from cross-sectional data. Group‐level obesity can be seen as an emergent property of a complex system, consisting of feedback loops between individual body weight perception, individual weight‐related behaviour and group‐level social norms (a product of group‐level ‘normal' body mass index (BMI) and socio-cultural ‘ideal' BMI). As overweight becomes normal, the norm might be counteracting health awareness in shaping individual weight‐related behaviour. System dynamics modelling facilitates understanding and simulating this system's emergent behaviour. Six system dynamics models (SDMs) were constructed based on an expert‐informed causal loop diagram and data from six socio-cultural groups (Dutch, Moroccan and South‐Asian Surinamese men and women). The SDMs served to explore the effect of three scenarios on group‐level BMI: ‘what if' weight‐related behaviour were driven by (1) health awareness, (2) norms or (3) a combination of the two. Median BMI decreased approximately 50% and 30% less in scenarios 2 and 3, respectively, than in 1. In men, the drop in BMI was approximately two times larger in scenario 1 versus 3, whereas in women, the drop was approximately equal in these scenarios. This study indicates that the overweight norm in men holds group‐level BMI close to overweight despite health awareness. Since norms are counteracting health awareness less strongly in women, other drivers of obesity must be more relevant. The SDMs were calibrated using a multi‐ethnic cohort data, which is a cross-sectional data; however, to study how the system evolves over time, longitudinal data is required. A method was developed to generate pseudo‐time series data from the available cross‐sectional data by generating a set of qualitative ‘data‐generating assumptions'. These assumptions are based on the system's temporal behaviour that is expected to exist across all groups. An example of such an assumption is that, on average, an individual can lose 2 kg/month. Here, linear dynamics was assumed for the system's short‐term behaviour. The time steps in the SDMs were assumed as months to identify an approximate timescale of the system's behaviour. However, this timescale is not exact, and therefore, the relative trends are more important to interpret than the exact timescale they occur on. Another method was developed for inferring computational models from cross-sectional data using Langevin dynamics. This method can be applied to any system that can be described as effectively following a free energy landscape which is stable and independent of any external force. A crucial assumption in this method is that the data-points are gathered from a system in (local) equilibrium. The result is a set of stochastic differential equations which capture the temporal dynamics, by assuming that groups of data-points are subject to the same free energy landscape and amount of noise. This is a `baseline' method which initiates the development of computational models which can be iteratively enhanced through the inclusion of expert knowledge. The proposed method can only estimate directions of progression, not velocities. Hence, the timescale of the predicted dynamics remains unknown. This timescale can be estimated from the data or from known statistical properties of the rates of change in reality; for instance, the fact that the maximum sustainable rate of weight loss observed in a population is about 2 kg per month. This method showed significant predictive power when compared against two population-based longitudinal datasets. The proposed method can facilitate the use of cross-sectional datasets to obtain an estimate of the underlying dynamics of the respective processes. The metabolic disorder of T2D is characterized by insulin resistance, beta-cell dysfunction, and apoptosis. Chronic hyperglycemia causes deterioration of beta-cell function through oxidative stress, endoplasmic reticulum (ER) stress, and cytokines. In this thesis, two computational models are proposed to study beta-cell dysfunction in T2D. The first computational model is a Boolean network model integrating the insulin resistance pathway with the beta-cell apoptosis pathway. This model has five input signals, namely, ER stress, oxidative stress, tumor necrosis factor alpha (TNF-alpha), Fas ligand (FasL), and interleukin-6 (IL-6). Simulations were performed using random order asynchronous update and with different combinations of the input signals. The model simulations were able to reproduce the expression levels of genes in T2D as reported in the literature. This model can be useful in studying the qualitative behaviour of important genes in the presence of oxidative stress, ER stress, and pro-inflammatory cytokines. Compromised beta-cell identity is emerging as an important contributor of beta-cell dysfunction in T2D. Most evidence suggests that this identity loss results from hyperglycemia-induced inactivation of transcription factors involved in mature beta-cell identity. Beta-cells with compromised identity gradually become dysfunctional with defective insulin secretion in response to glucose. An integrated mathematical model was developed to study the underlying mechanisms that regulate two important beta-cell identity transcription factors and regulators of insulin promoter activity, PDX1 and MAFA, and lead to their downregulation in the presence of chronic hyperglycemia. The aim of this work was to investigate the loss of beta-cell function through loss of beta-cell identity in the presence of chronic hyperglycemia. This model was used to study the changes in PDX1, MAFA and insulin mRNA levels under the effect of different glucose concentrations. In addition, this model was used to analyse the effect of different inhibitors of PDX1 and MAFA on these transcription factors and insulin mRNA levels. This integrated model can be a useful tool to further extend our understanding of the mechanisms leading to compromised beta-cell identity and beta-cell dysfunction in the presence of chronic hyperglycemia and identify potential intervention targets. Overall, the three studies of the social and biological factors of T2D demonstrate the importance of computational models in understanding the complex systems involved in the pathogenesis of T2D and designing effective intervention strategies. These computational models facilitate the evaluation of hypothetical scenarios in silico and simulation of the effect of interventions. This is especially advantageous for systems for which comparing counterfactual scenarios would not be possible or would be difficult in vivo.||URI:||https://hdl.handle.net/10356/155229||DOI:||10.32657/10356/155229||Rights:||This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).||Fulltext Permission:||embargo_20230301||Fulltext Availability:||With Fulltext|
|Appears in Collections:||IGS Theses|
Files in This Item:
|9.21 MB||Adobe PDF||Under embargo until Mar 01, 2023|
Updated on Jul 4, 2022
Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.