School of Scientific Visualization

From the 15th to the 21st of January 2017, the School of Scientific Visualization took place at the Cineca venue in Rome, within the Theoretical Chemistry and Computational Modeling (TCCM) European Joint Doctorate (EJD) program. The course was edited by the nonacademic partner SCAI Cineca together with Université Paul Sabatier in France and represents the second activity among the four mandatory activities planned for the second Ph.D. year.

The aim of the school was to offer a general view of the available tools used for scientific data visualization

and graphic representation. Several professors and experts on scientific visualization and virtual reality showed, through theory and practical exercises, the use of various opensource software and program languages capable of creating high-quality images, movies, and scientific animations. For every researcher, it is indeed important to know how to handle different graphic editors, especially nowadays where the creation of animations, and images are widely used to promote a better dissemination of scientific knowledge.

The school began Monday morning, with some examples illustrating how images, animations, and movies make the daily work of a researcher easier and provide a greater insight into the complex physical phenomena. It is inconceivable to look at a complex protein structure only by using plain text information of the system, namely numbers and characters. A much deeper insight is given visualizing the structure with molecular graphic tools e.g. Pymol, Jmol or VMD. Subsequently, the standard workflows and pipelines typically used by vector graphics libraries were discussed, along with vector graphics primitives in one, two and three dimensions. In the afternoon, the lectures focused on the most important 2D and 3D rendering techniques for scientific data visualization, such as color map, contouring, marching squares, or isosurfaces.

In the second day, the morning session was devoted to an overview of common programming languages and libraries used to increase the possibilities of chemical visualization tools. Afterwards the Chemical Markup Language (CML), developed in 1995 was introduced as one of the first library able to represent chemical information such as molecules, spectra, and crystallography data. The newer Q5 Cost project was also introduced. This library was born to overcome the lack of standards for data representation for quantum chemistry, defining a common data format capable of being shared by different quantum mechanical programs, such as Gaussian, Molcas, Molpro. The afternoon hands-on session focused on the Blender software, as an open-source tool for the creation of high-quality pictures and movie. The practical exercises afterwards were a great opportunity to further deepen the core concepts seen during the previous session.

The third day was mostly devoted to the Python programming language and its libraries for numerical computation and data analysis. After a general overview of the basic Python functions, the lectures focused mainly on the Numpy and Pandas libraries. The first one offers support for large multidimensional arrays and matrices, whereas the second provides data structures for numerical tables manipulation. During the afternoon, a series of practical exercises helped the attendants to gain insight into the practical implementation of Python scripts. The students learned how to use the two libraries and to handle complex data structures. The last lecture of the day was a tutorial dedicated to the PyMol molecular visualization system. This tool can produce high-quality 3D images of either small molecules and biological macromolecules like proteins. Furthermore, it is one of a few open-source visualization programs available for use in structural biology. The PyMol graphical interface and the most important features of the program were presented, highlighting the different types of render modes for complex protein structures.

Fourth day’s morning session was entirely dedicated to the visualization of chemical structures. Many programs nowadays offer several types of visual styles, such as wireframe, ball, and stick, spacefill, ribbon or surface. Each one highlights some important features of the system. The selection of one or another representation depends upon the information that the picture has to transmit to the reader. The afternoon hands-on session focused on the Jmol software. Thanks to a series of exercises the students experienced various features of the program and learned how to use the Jmol interactive scripting console, in order to automate different tasks of the software.

During the last day, lectures focused on classical dynamic simulations and creation of movies starting from molecular simulation. At the end, the VMD software suite and its features were presented.

The organizers of the school also created a contest for the Best Graphic Award. Each student had to submit a scientific oriented picture, movie or animation. The best three works will be awarded during the 2nd Annual Meeting in July 2017. Winners will get for free thousands of CPU hours on the High Performance Computing Marconi supercomputer, one of the world’s fastest supercomputers, owned by Cineca.

During the whole week, the 15 Ph.D. students had also the possibility to enjoy charm, grace, and beauty of the ethereal city. Furthermore, every evening dinner took place in different restaurants around the city. In this way, the students together with the organizers enjoyed good pasta, pizza, porchetta, and wine.

Francesco Talotta

 

Download this article as a pdf file!

Leave a Reply

Your email address will not be published. Required fields are marked *