A gesture-based interface for the exploration and classification of protein binding cavities

Abstract:

Molecular biologists seek to explain why similar proteins bind different molecular partners. The visual examination and comparison of binding cavities in protein structures can reveal information about the molecular partners that bind to a protein. Similarities can reveal regions that accommodate similar molecular fragments, while differences in binding preferences can arise from regions where binding sites differ. By comparing the binding cavities of multiple proteins, further information about the specificity of each protein can be discovered. But the visual examination of protein structure is a difficult cognitive task that requires persistence and quantitative precision. Software supports these efforts, but software for analyzing structure is difficult to use for investigators without computational backgrounds. By enabling non-computational users to better use analytical software, we hope to support progress in structural biology. Below, we present LeapRenderer, a three dimensional gesture-driven interface for visualizing protein surface models controlled primarily by the Leap Motion Controller. LeapRenderer serves exploration and classification functions. It allows researchers to explore the protein structures by manipulating 3D renderings via rotation and scaling. It also aids researchers in categorizing similar proteins into groups by providing a simple interface for comparing and sorting protein surfaces. These capabilities thereby support the discovery and classification of protein binding sites.