Abstract: The crowd sourcing is becoming a popular solution to big data problems, and virtual
community is gradually applied to scientific research. People also begin to pursue the high-quality
rendering as well the universal rendering methods. This paper studies a protein macromolecules 3D
visualization method, which is to manage the large number of atoms by the particle system, so as to
reduce the rendering time of each atom by improving the coloration method, thus it can express
Licorice and Van der Waals model, the secondary structure and some other conventional expressions.
Taking molecules from the 103 to105 scale as application examples, comparison of the static and
interactive efficiency between the particle system method and the popular method shows that the
former has better performances in both aspects. The domain decomposition of the molecular
structure is optimized, and the interactive performance is improved apparently. The results can run
smoothly when being published to the iPhone

Key words: molecular visualization, virtual science community, virtual world, big data;
particle system, conventional expression, OpenGL es2.0