Emerging Technologies in Scientific Data Visualisation
Location: Scuola Normale Superiore CECAM-IT-SISSA-SNS Node
As the saying goes, “a picture is worth a thousand words”, visualisation allows us to tap into high-bandwidth cognitive hierarchies of our brains and allows us to process high densities of information at once.
In the field of atomistic and molecular simulations, visualisation is a key element to research: we use ball-and-stick figures to represent the simulation scenarios, graphs to recognize and communicate parametric relationships of equations; we overlay the numerical results of simulations on atomistic models to discover new structural relationships and we extend our understanding of the atomistic world by using perceptual inferences, as in the case of protein folding, that are easy for humans yet difficult for computers.
Currently, the world of data and visualization is changing rapidly. In computational material science, the “Big Data” trend gave rise to several projects with vast output of data, many data-driven approaches are being introduced (see upcoming CECAM Workshop: "Big-Data driven Materials Science"). For instance, a new EU Center of Excellence, "NOMAD", is established to collect, store and regularize data to build a materials encyclopedia. As success stories of data-driven research emerge, e.g. the successful prediction of dielectric breakdown using machine learning, data as a raw material can be expected to attract more attention. While many aspects of the data production, analysis, and storage are being tackled, visualisation sits at the crossroad of scientific discovery and technological innovations.
The kind, size and accessibility of data being produced is fundamental to the visualization strategies to be developed. For example, atomistic visualizers such as VMD have adopted GPU acceleration to visualize large biomolecules. With the advance of quantum mechanical simulation tools, the representations of large biomolecules now contain more of volumetric and/or time-varying data. A recent example of the increased complexity of atomistic data visualization is a movie, describing the exciton coupling with atomic motion in the photosynthetic apparatus of bacteria, achieved with a multiscale visualization. Another example is the Time Dependent Density Functional Theory analysis of the ultrafast photoelectron transfer process in dye-sensitized solar cells, a study showing how data and visualization are coupled and how visual analytics contributes to the understanding of data besides communication of it.
Recently visual analytics is making its way into material simulations in novel ways. Some notable examples are i) a successful crystal structure prediction study using data clustering method supported by visual analytics, ii) a time-aggregated 2D heat-map method that reduces the time to explore inner tunnels of proteins, iii) a study where visual and haptic feedback was coupled with molecular dynamics to explore the low energy pathways of reactions, iv) immersive molecular viewers that utilize virtual reality technology allowing us to comprehend spatial organization of molecules as never before.
However, despite its great potential, visual analytics beyond ball-and-stick representation of molecules is still an emerging field. Several aspects are still waiting to be identified, discussed and developed in a collaboration between the different communities that contribute to scientific visualization.
Modern atomistic and molecular sciences make extensive use of computational methods and simulations to model and predict the properties of a great variety of systems. With the growing computational power available, following the “Big Data” trend, the results of simulations produce data sets of increasing size and complexity. While this situation creates a necessity for smarter data analysis procedures, it does not diminish the need for human intervention; on the contrary, it invites us to take better advantage of the high-bandwidth and parallel-processing capability of our brains using visual analytics.
Visual analytics exploits the strengths of visual cognition in quickly identifying structures, patterns, and anomalies in an image; and stimulates the mental processes for insight and understanding of large data sets and/or complex phenomena. Besides charts, in computational material science we refer to visual analytics regularly to understand the spatial organization of atomic simulation environments using graphical viewer and modeller software.
Recently visual analytics is making its way into material simulations beyond traditional ways. Some examples are visually-supervised data clustering during crystal structure prediction and heat-maps to explore the voids inside proteins. Technological advancements are also promising new ways of interacting with data such as visual and haptic guidance of molecular dynamics to explore low energy pathways of reactions, and immersive virtual reality experiences that finally allow to grasp the 3D arrangement of complex molecules at a glance.
However these are just early examples; and, despite its great potential, visual analytics beyond XY plots or ball-and-stick representations is an emerging field. Several aspects are still to be identified and discussed between different communities that contribute to scientific visualization and to be developed in collaboration. Open questions of the state-of-the-art are the following:
-Data Producers: What are the emerging visualization needs for Big Data; how are they different than scaled-up versions of existing tools?
-Data Analysts: How to enhance current analysis tools or create new ones with visualization? What visual analytics techniques, representations and mapping methods can we borrow from other fields now that the molecular simulations can produce a variety of data other than molecular representations?
-Technologists: How can we better use the developing technologies such as Virtual Reality, haptic feedback mechanisms, graphical artificial neural networks, and computer vision to reveal patterns and relationships that were previously not exposed to visualization at all? How can computational material science benefit from technologies that target interaction with data?
In this workshop we bring three communities together, data producers, analysts and visualization technologists to build a knowledge network to answer these questions. The workshop program is complemented with VR demo sessions for the participants to try new visualization technologies and exchange hands-on information and experience. A round table session is planned in the second day to establish a shared document in which open questions and possible future directions to answer them will be listed in the form of open challenges.
Emine Kucukbenli (SISSA) - Organiser
Giordano Mancini (Scuola Normale Superiore) - Organiser
Monica Sanna (Scuola Normale Superiore) - Organiser
Stefano de Gironcoli (Scuola Internazionale Superiore di Studi Avanzati - International School for Advanced Studies) - Organiser