Department of Medical Informatics and Clinical Epidemiology
Oregon Health & Science University
Abstract Genotypes and phenotypes are fundamental to many areas of the life sciences. Biomedical and animal model databases contain a wealth of information about genetics, genomics, embryonic development, phenotypes, and diseases, presenting biomedical researchers the opportunity to uncover human disease mechanisms. Currently, however, investigators lack suitable informatics tools to eectively meet the challenge of traversing these rich datasets, synthesizing the biology described therein, and identifying gaps in our knowledge that need further research. In this study, I address the question of how a visual analytics tool can help scientists investigate these important biological relationships. I developed a prototype of a tool that enables users to visually explore connections between human and animal model data rendered as a network graph. Two iterations of a prototype were produced, including mockups and a functional software-based application. A small group of users was recruited to provide feedback, identify needs, and evaluate the prototype's usability. With different user interface design and evaluation approaches from which to choose and a brief time frame in which to conduct the study, I employed a user-centered design philosophy, qualitative "discount" evaluation methods, and software libraries wellvii suited to making a visually appealing and highly responsive interface. The study generated abundant high-quality user feedback, a generally usable early software prototype, and a comprehensive set of recommendations for subsequent tool development, confirming the effectiveness of these choices.
School of Medicine
Segerdell, Erik Jon, "How can an interactive visual analytics tool help biomedical scientists investigate genotype-phenotype relationships?" (2015). Scholar Archive. 3736.
Available for download on Sunday, December 09, 2018