Visualization of molecules, and their behavior over time helps in scientific exploration, discovery and a means of confirming what we know about them. Simulations are performed at the molecular level with the increase in resolution of data sets and parallel computing machines. At CVC, we have many visualization tools developed for interactive browsing of large data sets.

Proteins, made up of atoms, are inherently non planar shapes. There are many models which are used to represent them, including

• Ball and stick model
• Union of balls model
• Solvent accessible surface
• Solvent contact surface
• Rolling solvent model

Apart from surfaces, we are also interested in visualization of fields including interactions, electron density and electrostatic potentials. These volumetric functions are visualized in a variety of software tools developed at CCV, which include ray tracing, hardware based texture mapped rendering and hierarchical splatting.

Some of the visualization methods we have to represent and render large data sets are

• Imposter model for large ball and stick models
  

  Close up view of the microtubule	Zoomed out view of the microtubule, rendered on a PC at interactive speeds

• Splatting of AMR and higher dimensional data sets
  

  5D visualization of interactions

• Volumetric visualization based on hardware accelerated texture mapping algorithms
  
  
  
• Multiresolution based isocontouring

• Volumetric ray tracing
  
  

• Algbraic functional representation for curves and surfaces of molecules
  
  

• Client server rover based model for large data exploration
  This is useful to visualize very large models of volumetric functions, allowing the user to explore regions of interest in high resolution, guided by a low resolution rendering of the same data set in a different synchronized view.