Molecular Dynamics Simulations (MD) plays an essential role in the field of computational biology. The simulations produce extensive high-dimensional, spatio-temporal data describ-ing the motion of atoms and molecules. A central challenge in the field is the extraction and visualization of useful behavioral patterns from these simulations. Throughout this thesis, I collaborated with a computational biologist who works on Molecular Dynamics (MD) Simu-lation data. For the sake of exploration, I was provided with a large and complex membrane simulation. I contributed solutions to his data challenges by developing a set of novel visual-ization tools to help him get a better understanding of his simulation data. I employed both scientific and information visualization, and applied concepts of abstraction and dimensions projection in the proposed solutions. The first solution enables the user to interactively fil-ter and highlight dynamic and complex trajectory constituted by motions of molecules. The molecular dynamic trajectories are identified based on path length, edge length, curvature, and normalized curvature, and their combinations. The tool exploits new interactive visual-ization techniques and provides a combination of 2D-3D path rendering in a dual dimension representation to highlight differences arising from the 2D projection on a plane. The sec-ond solution introduces a novel abstract interaction space for Protein-Lipid interaction. The proposed solution addresses the challenge of visualizing complex, time-dependent interactions between protein and lipid molecules. It also proposes a fast GPU-based implementation that maps lipid-constituents involved in the interaction onto the abstract protein interaction space. I also introduced two abstract level-of-detail (LoD) representations with six levels of detail for lipid molecules and protein interaction. Finally, I proposed a novel framework consisting of four linked views: A time-dependent 3D view, a novel hybrid view, a clustering timeline, and a details-on-demand window. The framework exploits abstraction and projection to enable the user to study the molecular interaction and the behavior of the protein-protein interaction and clusters. I introduced a selection of visual designs to convey the behavior of protein-lipid interaction and protein-protein interaction through a unified coordinate system. Abstraction is used to present proteins in hybrid 2D space, and a projected tiled space is used to present both Protein-Lipid Interaction (PLI) and Protein-Protein Interaction (PPI) at the particle level in a heat-map style visual design. Glyphs are used to represent PPI at the molecular level. I coupled visually separable visual designs in a unified coordinate space. The result lets the user study both PLI and PPI separately, or together in a unified visual analysis framework.
Rigid Docking Based Protein-Protein Interaction Prediction using High Scoring Docking Models