Computer simulation of water molecules

Computer Simulation of Water Molecules at Mineral Surfaces... 

An analysis of computer simulations of water at different pressures by Hummer et al. (110) suggested that hydrophobic contact pairs become increasingly destabilized with increasing pressure. The proposed scenario could explain the pressure denaturation of proteins as a swelling in terms of water molecules that enter the hydrophobic core by creating water-separated hydrophobic contacts. Additional support for the validity of Hummer s IT-model analysis has been achieved by pressure-dependent computer simulation studies of isolated pairs of hydrophobic particles, as well as rather concentrated solutions of hydrophobic particles (111, 112). Recently, the pressure-induced swelling of a polymer composed of apolar particles at low temperatures can be observed (113). 

The experimental and computational study of bacterial thioredoxin, an E. coli protein, at THz frequencies is presented. The absorption spectrum of the entire protein in water was studied numerically in the terahertz range (0.1 - 2 THz). In our work, the initial X-ray molecular structure of thioredoxin was optimized using the molecular dynamical (MD) simulations at room temperature and atmospheric pressure. The effect of a liquid content of a bacterial cell was taken into account explicitly via the simulation of water molecules using the TIP3P water model. Using atomic trajectories from the room-temperature MD simulations, thioredoxin s THz vibrational spectrum and the absorption coefficient were calculated in a quasi harmonic approximation. 

Over the past 15 years, a large number of papers have appeared dealing with computer simulations of water structure, thanks to the increased capacity and availability of fast computers. Simulations are based on accepting a reasonable expression for the pairwise interaction of water molecules, namely, the pair potential (energy) function. Much has been learned from simulation studies, and the results are valuable hints as to what the structure of liquid water may be, but not necessarily as to what the structure must be. Because of computational limitations, the typical sample of water molecules used in such analyses is about 500. If these molecules formed a small droplet, the radius would be five molecules, half of which would be from the outside layer. Thus, one must expect some dramatic surface effects. Furthermore, as discussed below, the calculations are based on a pairwise potential energy function this is probably a most serious limitation, as also discussed below. 

As discussed by Finney, the near trigonality of the charge distribution in the water molecule has also been taken into account in attempts to construct more realistic potential functions in molecular dynamics (MD) computer simulations of water and aqueous solutions. 

Thus, we have found unexpected complexities and even in this simple system have not yet been unable to accurately extrapolate the results of simulations done over periods varying from 1 to several hundred ps, to the low-friction conditions of extraction experiments performed in times on the oi dc r of ms. The present results indicate that one should not expect agreement between extraction experiments and simulations in more complex situations typically found in experiments, involving also a reverse flow of water molecules to fill the site being evacuated by the ligand, unless the simulation times are prolonged well beyond the scope of current computational resources, and thereby strengthen the conclusion reached in the second theoretical study of extraction of biotin from it.s complex with avidin [19]. 

The explicit definition of water molecules seems to be the best way to represent the bulk properties of the solvent correctly. If only a thin layer of explicitly defined solvent molecules is used (due to hmited computational resources), difficulties may rise to reproduce the bulk behavior of water, especially near the border with the vacuum. Even with the definition of a full solvent environment the results depend on the model used for this purpose. In the relative simple case of TIP3P and SPC, which are widely and successfully used, the atoms of the water molecule have fixed charges and fixed relative orientation. Even without internal motions and the charge polarization ability, TIP3P reproduces the bulk properties of water quite well. For a further discussion of other available solvent models, readers are referred to Chapter VII, Section 1.3.2 of the Handbook. Unfortunately, the more sophisticated the water models are (to reproduce the physical properties and thermodynamics of this outstanding solvent correctly), the more impractical they are for being used within molecular dynamics simulations. 

There are cases in which one is interested in the motion of a biomolecule but wishes also to study the effect of different solvent environments on this motion. In other cases, one may be interested in studying the motion of one part of the protein (e.g., a side chain or a loop) as moving in a solvent bath provided by the remainder of the protein. One way to deal with these issues is, of course, to explicitly include all the additional components in the simulation (explicit water molecules, the whole protein, etc.). This solution is computationally very expensive, because much work is done on parts of the system that are of no direct interest to the study. 

Modem understanding of the hydrophobic effect attributes it primarily to a decrease in the number of hydrogen bonds that can be achieved by the water molecules when they are near a nonpolar surface. This view is confirmed by computer simulations of nonpolar solutes in water [15]. To a first approximation, the magnimde of the free energy associated with the nonpolar contribution can thus be considered to be proportional to the number of solvent molecules in the first solvation shell. This idea leads to a convenient and attractive approximation that is used extensively in biophysical applications [9,16-18]. It consists in assuming that the nonpolar free energy contribution is directly related to the SASA [9],... 

Run the dynamics for 1000 iterations or until you are satisfied that the run is producing a relatively constant set of attributes. Repeat the run 10 times and compute the average values for each attribute. Convert each value to a fraction of the total number of water molecules, 2100 in this case. The sum of the fx values then equals 1.0. These fractions become a structural profile of water at any simulated temperature. 

The advantages of the simple approach outlined above are the limited number of water molecules needed in the simulation and the well-defined water structure. The major drawback is that, owing to the periodicity, this water structure fits best on a (111) or (lll)-like surface, e.g., (211). There are at least two other approximations to model the water interaction. One is to include a large number of water molecules and apply molecular dynamics to determine a structure for the water and include this water arrangement in the simulations [Filhol and Neurock, 2006]. The drawbacks of this approach are the computational time required and the results sensitivity to the water structure. 

The effect of water molecules on pericyclic reactions can also be compared with the effects of Lewis acids on these reactions. The enhanced polarization of the transition state in these reactions would lead to stronger hydrogen bonds at the polar groups of the reactants, which will result in a substantial stabilization of the transition states in the same way Lewis acids do. A computer-simulation study on the Diels-Alder reaction of cyclopentadiene by Jorgensen indicated that this effect contributes about a factor of 10 to the rates.7... 

Garofalini, S.H. (1990) Molecular dynamics computer simulations of silica surface structure and adsorption of water molecules, J. Non-Cryst. Solids, 120, 1. 

The experimental results mentioned above as well as computer simulation data confirm the earlier concept of Bockris et about the reorientation of water molecules with a change of electrode potential. 

The calculations of the stmcture of water between charged flat walls show that the density profile becomes asymmetric and that there is enhanced structuring. This enhanced structuring is intimately connected with the possibility of a continuous phase transition in quasi two-dimensional systems, a subject of recent intense interest. ° Most of the molecular dynamics computer simulations on the effects of an external field have been carried out in an attempt to clarify the field-induced restructuring of water molecules at the metal surface, for which recent experimental data have become available. ... 

The required computational effort for a MD study is governed by various elements. Foremost the number of particles N is a. crucial factor, as the number of interactions is proportional to N2 or even higher (N3 ), if quantum chemical methods are applied as it is the case in CPMD simulations. In the present CPMD simulations the number of water molecules employed ranges between 60 and 90, which are treated on GGA density functional level. In the context of QMCF simulation studies of hydrated systems a solute and up to 50 solvent molecules treated by ab initio quantum mechanics are surrounded by 500-1000 water MM molecules to ensure that a sufficient number of bulk molecules is included. 

---