Time-series data can be difficult to analyse and interpret
Insulin is a key ingredient in cell metabolism and plays an essential role in long-term health, in obesity, and in diseases such as diabetes.
When we eat glucose it comes into the bloodstream and triggers insulin secretion. Insulin then travels via the blood and binds to receptor molecules, principally on muscle, fat or liver tissue. The insulin binds to that tissue, which initiates a convoluted series of thousands of molecular events.
Recent research to measure the complex processes that take place inside a cell stimulated by insulin created a new kind of dataset. However, there were no adequate methods or tools to analyse, visualise and interpret the multidimensional datasets that resulted from these studies.
Visualising data to understand complexity
Working in an interdisciplinary team with colleagues from the University of Sydney on the BioCode project, we set about explaining the complexities of the insulin signaling pathway to make it easier for researchers to understand and interpret the data they had collected. We did this by applying data visualisation, integration, user experience design and graphic design to the research problem.
Showing how insulin works with Minardo
A technical diagram with a large u-shape with purple centre representing a cell's nucleus; around the outside are numerous lines and arrows indicating the process of insulin signaling inside the cell.
Our data visualisation team created Minardo, a web-based tool inspired by Charles Minard, a nineteenth Century French civil engineer who was a pioneer in the field of information graphics.
Minardo condenses multiple dimensions of information, including time and a range of specific biochemical processes, into one image. An interactive online version allows users to see moving parts and find additional information by hovering over the displayed events.
As a result of this simplification, we helped improve understanding of a vitally important but poorly understood process. The set of methods and tools we developed can also be used by other life scientists to understand similar complex datasets.