Sampling, multiple molecular dynamics method

Woo TK, Margl P, Blochl PE, Ziegler T (2002) Sampling phase space by a combined QM/MM ab initio Car-Parrinello molecular dynamics method with different (multiple) time steps in the quantum mechanical (QM) and molecular mechanical (MM) domains. J Phys Chem A 106 1173-1182... 

In such cases, the MEHMC method could be employed in combination with an enhanced sampling method that deforms the effective energy surface (but preserves the location of the potential minima), such as that in [29, 97]. Likewise, it may be worthwhile to explore the use of a reversible multiple-time-scale molecular dynamics propagator [103] with MEHMC to accelerate the dynamical propagation. 

One of the ways of circumventing the problem of finding multiple energy minima of complex molecules is to turn to more sophisticated techniques that are capable of sampling phase space efficiently without the need to home in on particular minimum energy conformations. The two most useful techniques are molecular dynamics (MD) and the Monte Carlo (MC) method. Both approaches make use of the same types of potential functions used in molecular mechanics, but are designed to sample conformation space such that a Boltzmann distribution of states is generated. MC and MD techniques for molecular systems have been widely reviewed [11-14], and only the basics of the two methods are described below. 

Molecular Dynamics and Hybrid Monte Carlo in Systems with Multiple Time Scales and Long-range Forces Reference System Propagator Algorithms). Other variations include modifications of the potential in MD or the probability of acceptance employed in the MC method. A sampling of the more successful algorithms is presented below. 

---