DL_POLY is a molecular dynamics code and as such, it is based on the concepts of point objects – atoms, and central forces acting between them. Implicit in this picture is an assumption that all atoms are isotropically shaped – ask someone to sketch you an atom and, highly likely, you will be presented with a spherical object. Molecules are different: they come in different shapes and denominations, which is what makes chemistry an exciting subject. Of course, DL_POLY can handle molecules by treating them as assemblies of atoms connected by (chemical) bonds. However, when the object of study is a complex biological system containing many millions of atoms , a full atomistic description becomes inefficient and a coarse-grained description would provide an answer at substantially reduces costs . In certain cases, one level of coarse-graining is not sufficient and several levels are combined at different resolution to tackle complex problems . Development of multiscale models for complex chemical systems is now an integral part of theoretical research in life sciences and the seminal work of pioneers in this field, Martin Karplus, Michael Levitt and Arieh Warshel, has been acknowledged with a Nobel Prize in Chemistry in 2013 .
Figure 1. (A): Illustration of the fully atomistic and coarse-grained representation of 4-cyanobiphenyl molecules. (B): Schematic of the general loaded shape (with the centre of mass (COM) offset from the centre of force (COF), or centre of shape). (C): illustration of two shapes linked together by a bond (with the torsional angle shown) and the rest of the atomistic subsystem (broken lines).
We are extending the DL_POLY molecular dynamics package to include a general framework for shapes – a generic name for coarse-grained objects that cannot be described in classical pairwise central-force approach. Thus, Fig. 1(A) illustrates coarse-graining a molecule of mesogenic 4–cyanobiphenyl to a Gay–Berne  potential. Clearly, even in this simple case the interaction between two coarse-grained objects include orientation-dependent force. A more general case includes loaded shapes (see Fig. 1(B)) and shapes with additional interaction sites. Linking shapes together and with the atomistic subsystem by bonds provides the most general coarse-grained framework for biochemical and medical research.
- K. Y. Sanbonmatsu and C.-S. Tung, J. Struct. Biol. 157, 470 (2007).
- H. I. Ingólfsson et al., WIREs Comput Mol Sci 4, 225 (2014).
- A. Arkhipov Y. Yin and K. Schulten, Biophys J. 95, 2806 (2008).
- The Nobel Prize in Chemistry 2013 -- Advanced Information.
- G. Gay and B. J. Berne, J. Chem. Phys. 74, 3316 (1981).