SG-RAD: A Visual Analytics System in Subgroup and Risk Factors Analysis and Discovery

Subgroup studies can help identify risk factors that influence a detailed subpopulation of people and help provide insights into precision medicine. However, the complexity between patients of a specific disease and potential risk factors makes it hard for epidemiologists to find meaningful patterns. It is challenging for epidemiologists to go through all the possible different risk factors and analyze the relationships between risk factors. Therefore we developed an interactive visualization system called SG-RAD (SubGroup Risk factors Analysis and Discovery) to assist users in identifying subgroups, exploring said subgroup’s risk factors, and further investigating the relationship between them. Specifically, the system allows users to define the contrasting subgroup as well as the targeted variables, find notable patterns of risk factors and further investigate the risk factors’ influence within and outside the subgroup. We conduct a case study, to identify lifestyle habits that are considered as risk factors of gout in a subgroup within the same high-risk genetic group for gout in Taiwan Biobank’s population. Our system, SG-RAD, can assist our domain experts in finding patterns in a detailed group of people that can be considered as risk factors for disease and further investigate the risk factors influence within the subgroup.