Improving the usability of complex biological networks through interestingness measures and interactive visualization

Abstract

Complex biological networks are dynamic and intricate systems that reflect the fundamental processes of life. This can range from the molecular interactions in a single cell to the complex communication web between organs and tissues in multicellular organisms, where the network orchestrates different biological functions. Unravelling these complexities is critical to the development of therapeutic interventions. The interdisciplinary nature of studying biological networks involves integrating principles that cut across biochemistry, molecular biology, genetics, and system biology, thereby providing a comprehensive perspective that allows researchers to examine the intricate connections driving the complexities in living organisms. Biological systems are constantly changing, capturing, and understanding their dynamism can be challenging. Visualization tools can help solve these difficulties by simplifying the complexity and representing the data visually and intuitively, making it easier for the researchers to identify the inherent patterns and relationships within the data. The aim of this thesis is to examine the features and measures of interest and evaluate the usability of interactive visualization of complex biological networks. To achieve this, five objectives were formulated. First, tasks and patterns of interest regarding the analysis of biological networks were determined through a literature review, interviews, and consultations with biologists. Second, a set of metrics based on tasks and patterns was defined by demonstrating how the concepts of interestingness and visualization can support the analysis of complex biological networks. Third, the study evaluated the usability and limitations of existing network visualization methods used for biological networks to identify how the usability of complex biological networks can be improved. Fourth, a visualization tool was designed and developed that overcomes current limitations and supports human cognition and data exploration, using multiple coordinated views and interactivity guided by interestingness measures. Finally, the network visualization tool was evaluated to identify limitations and areas for improvement. Overall, the evaluation of the developed tool was positive and guided by experts’ feedback, which was obtained using survey and interview techniques.