Publications

Performance analysis is an integral part of developing and optimizing parallel applications for high performance computing (HPC) platforms. Hierarchical data from different sources is typically available to identify performance issues or anomalies. Some hierarchical data such as the calling context can be very large in terms of breadth and depth of the hierarchy. Classic tree visualizations quickly reach their limits in analyzing such hierarchies with the abundance of information to display. In this position paper, we identify the challenges commonly faced by the HPC community in visualizing hierarchical performance data, with a focus on calling context trees. Furthermore, we motivate and lay out the bases of a visualization that addresses some of these challenges.

We present MegaMol, a low-overhead prototyping framework for interactive visualization of large scientific data sets. We give a brief summary of related work for context and then focus on a comprehensive overview of the core architecture of the framework. This is followed by the existing and novel features and techniques in MegaMol that define its current functionality. MegaMol has originally been developed to support the visualization and analysis of particle-based data sets that, for instance, come from molecular dynamics simulations. Meanwhile, the software has evolved beyond that. New algorithms and techniques have been implemented to handle many diverse tasks, including information visualization. Additionally, improvements have been made on the software engineering side to make MegaMol more accessible for domain scientists, like an easy-to-handle scripting interface.

In this manuscript, we review the evolution of atomistic visualization over the last 12 years and put the development of the community in context with our own efforts within the DFG collaborative research center 716. The goal is to provide a comprehensive summary of all relevant work that has been conducted under the auspices of project D.3. In this project, we focused on the visualization and analysis of particle-based data sets, and on how to bring these visualizations onto the workstation of domain scientists without the need for a large rendering infrastructure. We discuss how our decisions and goals evolved over time and show the success stories and publications. Finally, we give an outlook on the challenges that still require additional research and to which extent the requirements and constraints of current research have changed the way visualization works after these 12 years.

We present a novel abstract visualization for the binding sites of proteins. Binding sites play an essential role in enzymatic reactions and are, thus, often investigated in structural biology. They are typically located within cavities. The shape and properties of the cavity influence whether and how easily a substrate can reach the active site where the reaction is triggered. Molecular surface visualizations can help to analyze the accessibility of binding sites, but are typically prone to visual clutter. Our novel abstract visualization shows the cavity containing the binding site as well as the surface region directly surrounding the cavity entrance in a simplified manner. The resulting visualization resembles a hat, where the brim depicts the surrounding surface region and the crown the cavity. Hence, we dubbed our abstraction Molecular Sombrero, using the Spanish term for 'hat'. Our abstraction is less cluttered than traditional molecular surface visualizations. It highlights important parameters, like cavity diameter, by mapping them to the shape of the sombrero. The visual abstraction also facilitates an easy side-by-side comparison of different data sets. We show the applicability of our Molecular Sombreros to different real-world use cases.

In this paper, we review the advances in molecular visualization over the last 12 years and put the development of the community in context with our own efforts in the DFG Collaborative Research Center (CRC) 716. This includes advances in the field of molecular surface computation and rendering, interactive extraction of protein cavities, and comparative visualization for biomolecules. Our main focus was on the development of methods that assist the interactive and explorative visual analysis of large, dynamic molecular data sets on single desktop computers. To meet this goal, we often developed GPU-accelerated algorithms, which is in line with the general research direction of the field. Over the last years, we made seminal contributions to the field of molecular visualization, which partially still constitute the state of the art development or provided the basis for follow-up works

This project aims at investigating droplet-related phenomena, especially the influence of forces acting on the surface. To this end, we visualize interface deformation using two quantities: interface stretching and interface bending. This allows the visual analysis of droplet behavior and interface-related forces, such as the surface tension force and forces induced by Marangoni convection. For the latter, we show an application case for debugging and aiding the implementation of the Marangoni term in the solver Free Surface 3D (FS3D). Furthermore, we demonstrate the usefulness of our approach for prediction of droplet breakup.

Vector field topology is traditionally visualized by explicit extraction of graphical objects that represent the topological skeleton, including critical points, periodic orbits, and the invariant manifolds converging to those in forward or reverse time. In this work, we present implicit visualization of vector field topology by means of derived scalar fields. We obtain these fields by seeding a forward and a reverse streamline at each of their samples, and determining which critical point, periodic orbit, or solid boundary they converge to. This provides a segmentation of the domain with respect to regions of qualitatively similar streamline behavior, and thus its topological structure.We present an adaptive sampling strategy of the fields, together with an approach to determine their convergence with respect to streamline integration. Using 2D and 3D fields, we exemplify the utility and interpretation of our approach.