Bridging between data and knowledge: New ways to better understand Alzheimer's Disease and Type 2 Diabetes comorbidity

Abstract

The holistic understanding of biological phenomena enabled by systems biology forms the basis of current research. As opposed to the principle of classical biology where behavior of a biological system is explained by studying individual constituents, systems biology evaluates many constituents of a system simultaneously to explain how molecular processes influence higher level biological phenomena. The integration of knowledge acquired through published literature and data generated from direct experimentation is the most important feature of systems biology because it facilitates communication between these worlds and overcomes each other’s weaknesses. Today, it has profound applications in a wide range of disciplines such as biomarker identification, drug discovery, network analyses and disease-mechanism identification. <br /> In this thesis, using state-of-the-art frameworks and technologies of systems biology, we have performed a comparative analysis of disease-specific models to depict mechanism-centric comorbid association between Alzheimer’s Disease (AD) and Type 2 Diabetes Mellitus (T2DM). We achieved this through two different methodologies where literature-based findings were validated with publicly available data and vice-versa. The findings from our first methodology illustrate cross-talk between several signaling pathways which eventually manifest characteristic features of AD and T2DM. Our findings provide a wider and global overview of previously suggested comorbidity between these diseases. Moreover, we have explored putative beneficial and harmful effects induced by Metformin, an FDA-approved T2DM drug which is considered a candidate repurposing drug for AD. With our second methodology, we have identified four pleiotropic genes to be involved in pathophysiological events of both AD and T2DM. Interestingly, these genes did not fall into the category of well-known genes of both diseases, suggesting a new mechanistic route to the comorbid association. In addition to the work that explores AD-T2DM comorbidity, this thesis also focuses on a work that devises a new algorithm to enable quantification of disease mechanisms. The algorithm, named Candidate Mechanism Perturbation Algorithm (CMPA), was able to demonstrate that the intensity of impairment of causal mechanisms is different across spatial and temporal resolutions. Such an implementation opens up the possibility to generate a ranked and prioritized list of disease mechanisms. <br /> Lastly, this thesis endorses understanding of disease mechanisms and network analysis by providing robust and reproducible workflows. The works presented here introduce methodologies to bring new insights about comorbid diseases and score disease-associated mechanisms.