Premium
MBO(N)D: A multibody method for long‐time molecular dynamics simulations
Author(s) -
Chun Hon M.,
Padilla Carlos E.,
Chin Donovan N.,
Watanabe Masakatsu,
Karlov Valeri I.,
Alper Howard E.,
Soosaar Keto,
Blair Kim B.,
Becker Oren M.,
Caves Leo S. D.,
Nagle Robert,
Haney David N.,
Farmer Barry L.
Publication year - 2000
Publication title -
journal of computational chemistry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.907
H-Index - 188
eISSN - 1096-987X
pISSN - 0192-8651
DOI - 10.1002/(sici)1096-987x(200002)21:3<159::aid-jcc1>3.0.co;2-j
Subject(s) - molecular dynamics , anharmonicity , multibody system , computer science , flexibility (engineering) , motion (physics) , set (abstract data type) , dynamics (music) , algorithm , statistical physics , computational science , physics , computational chemistry , classical mechanics , chemistry , mathematics , statistics , quantum mechanics , artificial intelligence , acoustics , programming language
A modeling approach that can significantly speed up the dynamics simulation of large molecular systems is presented herein. A multigranular modeling approach, whereby different parts of the molecule are modeled at different levels of detail, is enabled by substructuring. Substructuring the molecular system is accomplished by collecting groups of atoms into rigid or flexible bodies. Body flexibility is modeled by a truncated set of body‐based modes. This approach allows for the elimination of the high‐frequency harmonic motion while capturing the low‐frequency anharmonic motion of interest. This results in the use of larger integration step sizes, substantially reducing the computational time required for a given dynamic simulation. The method also includes the use of a multiple time scale (MTS) integration scheme. Speed increases of 5‐ to 30‐fold over atomistic simulations have been realized in various applications of the method. © 2000 John Wiley & Sons, Inc. J Comput Chem 21: 159–184, 2000