The Force-Field

It is important to retain sight of the fact that force-field calculations are done on a molecular model . This model is assigned properties which reproduce experimental facts, but this does not 

As will be shown in what follows, current force-fields are some distance from this point of convergence today. Some of them have stiff force constants, and soft atoms, and some have the reverse. Thus changes, or errors, if you will, in one group of parameters can be compensated for with adjustment of another group. 

Until the last few years, most force-fields described in the literature were developed exclusively for the interpretation of vibrational spectra. Such force-fields are usually of limited use in the present context, because typically they neglect many or all of the van der Waals interactions, and were not parameterized with structure and energy (apart from the spacing of the vibrational levels) in mind. Recently Lifson (Lifson and Warshcl, 1968 Warshel and Lifson, 1970) and Boyd (Boyd, 1968 Shieh et al., 1969 Boyd etal., 1971 Chang et al., 1970) have each developed force-fields aimed at fitting simultaneously to vibrational spectra, structure, and energy 

The full description of the interactions in the system that are included in the simulations is called the force field. A typical potential function of the system features extremely simplified forms (for example, harmonic terms) for the various contributions  

Here the atoms in the system are numbered by i, j, k, l = 1. N. The distance between two atoms i, j is ry, q is the (partial) charge on an atom, 6 is the angle defined by the coordinates (i, j, k) of three consecutive atoms, and 4 is the dihedral angle defined by the positions of four consecutive atoms, e0 is the dielectric permittivity of vacuum, n is the dihedral multiplicity. The potential function, as given in equation (6), has many parameters that depend on the atoms involved. The first term accounts for Coulombic interactions. The second term is the Lennard-Jones interaction energy. It is composed of a strongly repulsive term and a van der Waals-like attractive term. The form of the repulsive term is chosen ad hoc and has the function of defining the size of the atom. The Ay coefficients are a function of the van der Waals radii of the 

The first and second terms in equation (6) are rather expensive in terms of computer power needed. They include the interaction of all atoms with all other 

---

Hardy s explanation that the small coefficients of friction observed under boundary lubrication conditions were due to the reduction in the force fields between the surfaces as a result of adsorbed films is undoubtedly correct in a general way. The explanation leaves much to be desired, however, and it is of interest to consider more detailed proposals as to the mechanism of boundary lubrication. 

For a very large number of variables, the question of storing the approximate Hessian or inverse Hessian F becomes important. Wavefunction optimization problems can have a very large number of variables, a million or more. Geometry optimization at the force field level can also have thousands of degrees of freedom. In these cases, the initial inverse Hessian is always taken to be diagonal or sparse, and it is best to store the... 

Pulay P and Meyer W 1971 Ab initio calculation of the force field of ethylene J. Moi. Spectrosc. 40 59... 

Atomistically detailed models account for all atoms. The force field contains additive contributions specified in tenns of bond lengtlis, bond angles, torsional angles and possible crosstenns. It also includes non-bonded contributions as tire sum of van der Waals interactions, often described by Lennard-Jones potentials, and Coulomb interactions. Atomistic simulations are successfully used to predict tire transport properties of small molecules in glassy polymers, to calculate elastic moduli and to study plastic defonnation and local motion in quasi-static simulations [fy7, ( ]. The atomistic models are also useful to interiDret scattering data [fyl] and NMR measurements [70] in tenns of local order. 

But the methods have not really changed. The Verlet algorithm to solve Newton s equations, introduced by Verlet in 1967 [7], and it s variants are still the most popular algorithms today, possibly because they are time-reversible and symplectic, but surely because they are simple. The force field description was then, and still is, a combination of Lennard-Jones and Coulombic terms, with (mostly) harmonic bonds and periodic dihedrals. Modern extensions have added many more parameters but only modestly more reliability. The now almost universal use of constraints for bonds (and sometimes bond angles) was already introduced in 1977 [8]. That polarisability would be necessary was realized then [9], but it is still not routinely implemented today. Long-range interactions are still troublesome, but the methods that now become popular date back to Ewald in 1921 [10] and Hockney and Eastwood in 1981 [11]. 

A number of issues need to be addressed before this method will become a routine tool applicable to problems as the conformational equilibrium of protein kinase. E.g. the accuracy of the force field, especially the combination of Poisson-Boltzmann forces and molecular mechanics force field, remains to be assessed. The energy surface for the opening of the two kinase domains in Pig. 2 indicates that intramolecular noncovalent energies are overestimated compared to the interaction with solvent. 

In our implementation of SMD, modified versions of VMD and Sigma communicate with each other using a customized, lightweight protocol. Sigma sends atomic positions resulting from each molecular dynamics time step to VMD for display. When the user specifies restraints on parts of the displayed model, VMD sends them to Sigma, where they are converted into potential-well restraints added to the force field [21]. 

Time-reversible energy conserving methods can be obtained by appropriate modifications to the (time-reversible) midpoint method. Two such modifications are (i) scaling of the force field by a scalar such that total energy... 

MTS methods exploit the existence of different time scales arising from the many interactions present in the force field. For expository purposes assume... 

The fifth and final chapter, on Parallel Force Field Evaluation, takes account of the fact that the bulk of CPU time spent in MD simulations is required for evaluation of the force field. In the first paper, BOARD and his coworkers present a comparison of the performance of various parallel implementations of Ewald and multipole summations together with recommendations for their application. The second paper, by Phillips et AL., addresses the special problems associated with the design of parallel MD programs. Conflicting issues that shape the design of such codes are identified and the use of features such as multiple threads and message-driven execution is described. The final paper, by Okunbor Murty, compares three force decomposition techniques (the checkerboard partitioning method. 

In stead, these m eth od s solve the poten tial energy surface by using a force field equation (see Molecular Mechanics" on page2] i.The force field equation represen ts electron ic energy implicitly th roil gh param eteri/ation. 

The force field ec uations for M.Vf+, AMBER, BlOg and OPES are similar in the types of terms they contain bond, angle, dihedral, van der Waals. and electrostatic. There are som e differences m the form s of the etinations that can al fect your ch oice of force field. 

MM+ is iiiiipne among the force fields in ihe way it treats bonds and angles. Both the bond and angle terms can contain higher... 

Hach molecular mechanics method has its own functional form MM+. AMBER, OPL.S, and BIO+. The functional form describes the an alytic form of each of th e term s in th e poteri tial. For exam pie, MM+h as both a quadratic and a cubic stretch term in th e poten tial whereas AMBER, OPES, and BIO+ have only c nadratic stretch term s, I h e functional form is referred to here as the force field. For exam pie, th e fun ction al form of a qu adratic stretch with force constant K, and equilibrium distance i q is ... 

In principle, atom types eoiild be assoeiated wilh a partieiilar parameter set rather than the functional form or force field. In HyperChern, however, atoms types are rigorously lied to a force field . M.M-t, AMBER, OPTS, and BIO+. Each of the force fields has a... 

Note All of the force fields provided in HyperChem are built on new irn picm en tatiori s of foree fields developed by various com pii-tational chemistry research groups. How-ever, HyperChem improves on the original force fields and uses new code. 

Although in teraetion s between vicinal I 4 atom s arc n om in ally treated as non bonded interactions, triost of the force fields treat these somewhat differently from normal 1 5 and greater non-bonded interactions. HyperCbern allows each of these 1 4 non-bonded interactions to be scaled down by a scale factor < 1.0 with AMBHR or OPI-S. bor HIO+ the electrostatic may be scaled and different param eters rn ay be ti sed for I 4 van dcr Waals interactions, fh e. AMBHR force field, for exam pie, n orrn a lly uses a seal in g factor of 0.5 for both van der Waals an d electrostatic interactions. 

The OPLS force field is described in twtt papers, one discussing parameters for proteins W. L. Jorgensen and J. Tirado-Rives,/. Amer. (. hem. Soc., 110, 1557 (iy8K) and on e discii ssin g param eters for n iicleotide bases [J. Pranata, S. Wiersch ke, and W. L. Jorgen sen. , /.. Amer. Chem. Soc.. 117, 281(1 ( 1991)1. The force field uses the united atom concept ftir many, but not all. hydrttgens attached to carbons to allow faster calculation s on macromolecular systems. The amino and nucleic acid residue templates in HyperChein automatically switch to a united atom representation where appropriate when th e OPLS option is selected. 

AtomTypeMass is set to the nam e of th e file that lists the atom types associated with the force field and their masses (masses are associated with a type h ere not an atomic n umber). Th e file can have any name hut by convention isnamed, for example, as amheiTyp.txtfdbn. 

V these three functional forms, the improper torsion definition is most widely used as it can then be easily included with the proper torsional terms in the force field. However, the... 

---