HiveArcs - Visualizing genome co-expression data

Abstract

The developing human brain is a complex process, governed by the human genome. Understanding this development process is the key to understanding not only the connections between genes and brain functioning, but can also advance knowledge about disorders such as Alzheimer’s and Autism. The BrainSpan transcriptional atlas, released in 2013 by the Allen Institute for Brain Science, provides unique opportunities to study this process, but also challenges existing methods for visual analysis. One particular problem arises when large co-expression networks are subdivided in modules through clustering algorithms. In this work we introduce HiveArcs, an interactive online visual analysis for genome co-expression data. HiveArcs visualizes modules using arc diagrams on radially placed axes. Using this lay-out accommodates display of per-gene tracks of additional information as well as relationships between modules. This allows for quick and intuitive comparison between modules as well as in depth analysis of a single module.