Scientific Data and Molecular Modeling

Introduction
In the field of computational chemistry, molecular models are created, numerical data are generated, and analyses are performed. Creative visualization of the numerical data (e.g. x, y, z coordinates, forces, correlation values and other derived quantities and observables) is essential for gaining insight during the analysis step. That is because the conformational characteristics of (marco-)molecules are easily revealed through visual inspection, which subsequently guide the statistical analysis. In that sense, the following statement strongly holds true in computational chemistry: “The purpose of visualization is insight not pictures. The main goals of this insight are discovery, decision-making, and explanation [1].”

Computational research generates large amounts of raw data and, consequently, more analytic output that requires visualization. Representing the same raw data in diverse ways can aid researchers in gaining new insights; simultaneously viewing these diverse analytic outputs can enable the observation of previously unseen correlations, or help create new ideas for performing further analysis. Visual analytics [2] has the potential to significantly benefit computational chemistry. This is particularly true for biological modeling, where the aspect of visualized data is arguably more important than in materials modeling (e.g. polymer research) since there are more atomistic building blocks, greater chemical diversity, more unique interactions, and more distinct conformations that govern macroscopic processes. Thus, even though both fields are related to each other, the methods of analysis do differ and matter.

Our research can be broadly categorized into the following topics:

  1. Visualization and interpretation of molecular dynamics data
  2. Molecular modeling on ultra-high resolution displays
  3. Haptic devices for molecular modeling
  4. Nonlinear presentation of scientific data

Example: Proposed Benefits of Ultra-high Resolution Displays for Natural Scientists
Researchers within the natural sciences can benefit from the simultaneous high-resolution display of multiple data visualizations. This can include data that is from different sources, or it could be the same data but visualized in different ways. Displaying several visualizations at high resolutions and at the same time increases the chances to observe patterns, correlations, artifacts, and rare events within the data. It can also confirm or disqualify the expectations that are developed during the course of conducting research. Exemplifying this in force-fields optimization would be the expectation that molecules with similar atomic frameworks will have similar rotational curves. However, the exceptions that do occur represent opportunities for revising one’s hypothesis and generating new ideas. Such anomalous data (i.e. artifacts) would be difficult to identify by statistical analysis alone, since the connection between cause and effect is lost.

An ultra high-resolution display system can also benefit researchers by facilitating collaborative discussions concerning data [3-6]. Clear and direct communication between collaborators is essential for productive discovery and knowledge generation, with visualizations often playing an essential role. As data sizes increase, and with visualizations gaining greater information content, their subsequent simultaneous presentations on large high-resolution displays enables efficient discussions to occur between researchers. This also allows for a more productive use of the researchers’ time since everyone is together and are viewing a large amount of results at the same time. Interdisciplinary collaborations have the potential to benefit the most from these large displays since it allows for the displaying of diverse data side-by-side and direct verbal communication of how that data may be interpreted within their respective fields.

Representative Publications

  1. Kirschner, K. N.; Reith, D.; Jato, O. & Hinkenjann, A. Visualizing potential energy curves and conformations on ultra high-resolution display walls Journal of Molecular Graphics and Modelling , 2015, 62, 174 – 180
  2. Krämer-Fuhrmann, O.; Neisius, J.; Gehlen, N.; Reith, D. & Kirschner, K. N. Wolf2Pack – Portal Based Atomistic Force-Field Development Journal of Chemical Information and Modeling, 2013, 53, 802-808
  3. Wolf, A. & Kirschner, K. N. Principal component and clustering analysis on molecular dynamics data of the ribosomal L11-23S subdomain Journal of Molecular Modeling, Springer-Verlag, 2013, 19, 539-549
  4. 4. Wolf, A.; Baumann, S.; Arndt, H.-D. & Kirschner, K. N. Influence of thiostrepton binding on the ribosomal GTPase associated region characterized by molecular dynamics simulation Bioorganic & Medicinal Chemistry, 2012, 20, 7194-7205

References
[1] S.K. Card, J.D. Mackinlay, B. Shneiderman, Readings in Information Visualization: Using Vision to Think, Interactive Technologies Series, Morgan Kaufmann, 1999.
[2] D.A. Keim, J. Kohlhammer, G. Ellis, F. Mansmann (Eds.), Mastering The Information Age – Solving Problems with Visual Analytics, Eurographics, 2010.
[3] S. Knudsen, M.R. Jakobsen, K. Hornbæk, An exploratory study of how abundant display space may support data analysis, in: Proceedings of the 7th Nordic Conference on Human-Computer Interaction: Making Sense Through Design, NordiCHI ’12, ACM, New York, NY, USA, 2012, pp. 558–567.
[4] T. Ni, G. Schmidt, O. Staadt, M. Livingston, R. Ball, R. May, A survey of large high-resolution display technologies, techniques, and applications, in: Virtual Reality Conference 2006, 2006, pp. 223–236.
[5] C. Andrews, A. Endert, B. Yost, C. North, Information visualization on large high-resolution displays: issues challenges and opportunities, Inf. Vis. 10 (4) (2011) 341–355.
[6] L. Renambot, B. Jeong, R. Jagodic, A. Johnson, J. Leigh, Collaborative visualization using high-resolution tiled displays, in: CHI 06 Workshop on Information Visualization and Interaction Techniques for Collaboration Across Multiple Displays, Montreal, Canada, 2006.