Peter L. Freddolino, Anton Arkhipov, Amy Y. Shih, Ying Yin, Zhongzhou Chen, and
Klaus Schulten.
Application of residue-based and shape-based coarse graining to
biomolecular simulations.
In Gregory A. Voth, editor, Coarse-Graining of Condensed Phase
and Biomolecular Systems, chapter 20, pp. 299-315. Chapman and Hall/CRC
Press, Taylor and Francis Group, 2008.
FRED2008
A wide variety of coarse-graining methods for biological systems
currently exist, ranging in some sense from united-atom models to
elastic network models. We focus on the principles and applications
of two classes of biological coarse-graining, namely residue-based
and shape-based coarse graining. Residue-based CG is a broad family
of methods in which clusters of 10-20 covalently bonded atoms are
represented by one bead; it is a fairly natural and common method
for coarse-graining when a speedup of 1-2 orders of magnitude over
all-atom simulations is required. Shape-based CG is a method recently
developed in our group which uses a neural network algorithm to assign
CG beads to domains of a protein, efficiently reproducing the shape
of the protein with a minimal number of particles. Interactions
between beads are then parameterized from all-atom simulations of
the bead components. In this chapter we present a summary of
both methods, along with exemplary applications of residue-based CG
to two lipid-protein systems involving large-scale conformational
changes, and of shape-based CG to the mechanical properties of
polymeric systems.