Visualization of Biological Pathways : An Ethnographic Study and Systems Evaluation

Life scientists use pathway diagrams to represent complex biochemical interactions at the molecular level. Although several visualization software systems have been developed to assist them in analyzing pathways, relatively few scientists use them extensively, and those that do rarely exploit their full capabilities. In collaboration with life scientists, we found that many are reluctant to use these systems due to the steep learning curve and amount of effort required to construct biologically relevant pathways. They find minimal value in a system that provides only simple visual or dynamic pictures, without providing adequate means to manipulate pathways in terms of analytic requirements. This work reports the results of ethnographic field studies conducted to understand life scientists’ needs for pathway analysis, and heuristic evaluations of five popular pathway visualization systems based on the identified data analysis requirements. These studies suggest several critical new requirements for pathway visualization systems that provide guidance for future development of these tools. The critical requirements include: 1) automated construction and updating of pathways by searching literature databases, 2) overlaying rich biological information on pathways in a biologically relevant format, 3) linking pathways to multidimensional data from high throughput experiments such as microarrays, 4) overviewing multiple pathways simultaneously and analysis of inter-connections between them, 5) scaling pathways to higher levels of abstraction to analyze effects of complex molecular interactions across levels of biological organization, and 6) logical organization and sharing of accumulated knowledge among researchers.