Force fields, limitations

MacKerell AD, Jr, M Feig, CL Brooks III (2004a) Extending the treatment of backbone energetics in protein force fields Limitations of gas-phase quantum mechanics in reproducing protein conformational distributions in molecular dynamics simulations. J. Comput. Chem. 25 (11) 1400-1415... 

Horinek, D. and R. R. Netz. 2011. Can simulations quantitatively predict peptide transfer free energies to urea solutions Thermodynamic concepts and force field limitations. Journal... 

This situation, despite the fact that reliability is increasing, is very undesirable. A considerable effort will be needed to revise the shape of the potential functions such that transferability is greatly enhanced and the number of atom types can be reduced. After all, there is only one type of carbon it has mass 12 and charge 6 and that is all that matters. What is obviously most needed is to incorporate essential many-body interactions in a proper way. In all present non-polarisable force fields many-body interactions are incorporated in an average way into pair-additive terms. In general, errors in one term are compensated by parameter adjustments in other terms, and the resulting force field is only valid for a limited range of environments. 

Much work remains to be done in the development of this approach to explore the advantages and limitations of the method. The method will be extended to force fields that include torsional terms large systems such as biological macromolecules will also be treated. 

N is the number of point charges within the molecule and Sq is the dielectric permittivity of the vacuum. This form is used especially in force fields like AMBER and CHARMM for proteins. As already mentioned, Coulombic 1,4-non-bonded interactions interfere with 1,4-torsional potentials and are therefore scaled (e.g., by 1 1.2 in AMBER). Please be aware that Coulombic interactions, unlike the bonded contributions to the PEF presented above, are not limited to a single molecule. If the system under consideration contains more than one molecule (like a peptide in a box of water), non-bonded interactions have to be calculated between the molecules, too. This principle also holds for the non-bonded van der Waals interactions, which are discussed in Section 7.2.3.6. 

Intensive use of cross-terms is important in force fields designed to predict vibrational spectra, whereas for the calculation of molecular structure only a limited set of cross-terms was found to be necessary. For the above-mentioned example, the coupling of bond-stretching (f and / and angle-bending (B) within a water molecule (see Figure 7-1.3, top left) can be calculated according to Eq. (30). 

Molecular mechanics force fields have much information built into them and can be accurate for the molecules used in their param eten/ation. For molecules outside the limited scope for 40. Dewar. J. S. Dicier, K. M../. Am. Chem. Soc. 108 807. ), 1086. 

In the case of ethylene, because of 2-fold symmetry, odd terms drop out of the series, V3, V5,... = 0. In the case of ethane, because of 3-fold symmeti-y, even temis drop out, V2, V4,... = 0. Terms higher than three, even though permitted by symmetry, are usually quite small and force fields can often be limited to three torsional terms. Like cubic and quaitic terms modifying the basic quadratic approximation for stretching and bending, terms in the Fourier expansion of Ftors (to) beyond n = 3 have limited use in special cases, for example, in problems involving octahedrally bound complexes. In most cases we are left with the simple expression... 

Molecular mechanics methods have only been used to a limited extent for these classes of compounds. However, molecular mechanics methods do fairly well in describing the geometries and relative energies of compounds with these elements. It is perhaps only for historical and economic reasons that molecular mechanics has not been used more for modeling these elements. Subsequently, there are not as many force fields available. 

The molecular structure input requires atom types to be assigned, which are not the same from one force field to the next. The input also includes a list of bonds in the molecule. There is not a module to automatically assign atom types. Most of the modules use a Cartesian coordinate molecular structure, except for a few that work with torsional space. The same keyword file is read by all the executables. A little bit of input is obtained by the program either interactively or from an ASCII file piped to standard input, which makes for a somewhat cryptic input file. This system of common input files and the user choosing which executables to run give TINKER the ability to run very sophisticated simulations while keeping the input required for simple calculations fairly minimal within the limitations mentioned here. 

This division is somewhat arbitrary siace it is really the pore size relative to the size of the sorbate molecule rather than the absolute pore size that governs the behavior. Nevertheless, the general concept is useful. In micropores (pores which are only slightly larger than the sorbate molecule) the molecule never escapes from the force field of the pore wall, even when ia the center of the pore. Such pores generally make a dominant contribution to the adsorptive capacity for molecules small enough to penetrate. Transport within these pores can be severely limited by steric effects, leading to molecular sieve behavior. 

It should also be noted that a force field for a wide variety of small molecules, CHARMm (note the small m, indicating the commercial version of the program and parameters), is available [39] and has been applied to protein simulations with limited success. Efforts are currently under way to extend the CHARMm small molecule force field to make the nonbonded parameters consistent with those of the CHARMM force fields, thereby allowing for a variety of small molecules to be included in computational smdies of biological systems. 

---