EVL PhD candidate Chihua Ma presents her preliminary research in visual analytics for biological networks.

Monday, June 13, 2016
12:30 PM - 2:00 PM
EVL Cyber-Commons, 2032 ERF

Abstract:
Biological systems are often modeled as networks in which vertices denote individual bio-units, and edges represent interactions between vertices, such as protein-protein interaction networks, gene regulatory networks, and neural networks. Biological networks feature spatial features, e.g. coordinates in the brain, and nonspatial features, e.g. the network structure. They change dynamically, e.g. brain activation. They are analyzed and explored at multiple levels. For example, brain activation data may range from the macroscale level to the microscale level. Last but not least, modeling these networks is often done over multiple examples or with different models.

This wealth of data, features, models, and potential hypotheses and experiments exceeds the analytical capabilities of machines. Visualization provides an effective way to help biologists understand, communicate, and gain insight into their biological data through visual analysis and exploration.

I have contributed to a survey of spatial-nonspatial integration in biology; designed several novel multi-scale visual encodings; and completed three applications that partially support the biological problems with the combination of those three challenges. In the future work, I will contribute to a taxonomy, design novel multi-scale visual representations and develop a visual approach that can fully support the big challenge.