[ Pobierz całość w formacie PDF ]

J
dx.doi.org/10.1021/ct500161f | J. Chem. Theory Comput. XXXX, XXX, XXX-XXX
However, considering the most general case of a complex
polyatomic charged ligand inserted into a charged
protein in solution, none of the schemes available at
present227,254,291-296 are sufficiently general, practical,
and accurate. Recently, a new method for removing finite
size effects has been proposed based on a continuum-
electrostatics analysis. It requires performing Poisson-
Boltzmann calculations on the protein-ligand system.178
The approach introduces the concept of the residual
integrated potential to account for the finite-size effect
related to the solvent-excluded volume of the protein and
the ligand, an effect that is absent in monatomic ion
solvation.
(6) Force-field inaccuracy. Due to the hundreds of
parameters involved in empirical force fields the
propagation of errors in these parameters on calculated
binding free energies is a complex problem. Recently,
Rocklin et al.297 investigated the sensitivity of binding
free-energy calculations to the nonbonded energy
parameters in force fieldsatomic radii, dispersion
well-depths, and partial chargesby performing tens of
thousands of small parameter perturbations. They
estimated that random, uncorrelated errors in force-
field nonbonded parameters must be smaller than 0.02 e
per charge, 0.06 Å per radius, and 0.04 kJ mol-1 per well
depth in order to obtain 68% confidence that a
computed binding affinity for a moderately sized lead
compound will fall within 4.2 kJ mol-1 of the true
affinity, if these are the only sources of error considered.
Fixed charge models of ligands, parametrized against
hydration free energies, might easily have larger
uncertainty in the partial charges, especially in nonpolar
binding sites.
7. CONFORMATIONAL CHANGES
The importance of knowing the change in free energy
associated with a change in molecular conformation was
already mentioned in the context of hydration or binding free
energy calculations. It relies on the definition of conformational
states as well as on the ability to define a reduced set of
(spatial) coordinates R(rN) on which the free energy is
projected. Such a hypersurface is commonly called a reaction
coordinate and, in configurational space, is a function of the
positions of atoms in the system. Note that the term reaction
coordinate is eventually associated with the minimum-free-
energy pathway connecting the reference state to the target
state but is commonly employed to characterize the order
parameter along which the variation of the free energy is
determined. The free energy as a function of the reaction
coordinate R′(rN), or the potential of mean force, is given by eq
5, where the term in curly brackets is the probability of finding
the system lying on the reaction coordinate. Difficulties related
to the representation of the reaction coordinate have been
sketched in the previous section. A comprehensive discussion
of methods to obtain reaction coordinates298-300 is beyond the
scope of the present review. Another difficulty is related to the
definition of conformational states. For small systems such as
carbohydrates, states can be relatively well-defined due to the
rigid nature of the glycosidic linkage. Carbohydrates are
therefore often used for testing new methodology301,302 or to
calibrate force fields.303 Small peptides in solution show
significantly more flexibility. Dipeptides are often used to test
Journal of Chemical Theory and Computation
Perspective
⃗
⃗
practical recommendations. Comparison of methods in
practical settings is therefore as important.257,329-332
(4) Interavailability of code. The transfer of a new method
developed, implemented and tested for one particular
MD software package to another one is far from being
trivial because today’s molecular dynamics packages are
very complex pieces of software developed over decades
often by a diverse group of contributors with different
backgrounds and experience.51 However, such transfer is
essential for a wider acceptance of a particular method
and also for a better comparison to other methods not
implemented in the original software. Plug-ins with
interfaces to different MD codes may help to disseminate
new methods to a wider community of users.333
(5) Critical use of experimental data. Experimental measure-
ments are invariably contaminated with errors, which
may affect the maximally possible correlation between
simulation and experiment that can be achieved.324,334,335
Often modelers try too hard to reproduce experimental
data as precisely as possible, ignoring the fact that these
data are also subject to uncertainty.334,336-338
Journal of Chemical Theory and Computation
Perspective
Corresponding Author
*Phone: +49 711 685-66112. Fax: +49 711 685-66140. Email:
hansen@itt.uni-stuttgart.de.
Notes
The authors declare no competing financial interest.
F., Weiner, P. K., Eds.; Escom Science Publishers: Leiden, The
Netherlands, 1989; pp 27-59.
(13) Frenkel, D. In Computer simulations in Materials Science; Meyer,
M., Pontikis, V., Eds.; NATO ASI Series E205; Kluwer Academic
Publishers: Dordrecht, 1991; pp 85-117.
(14) Straatsma, T. P.; McCammon, J. A. Annu. Rev. Phys. Chem.
1992, 43, 407-435.
(15) Mark, A. E. In Encyclopedia of Computational Chemistry; von
A., Schaefer, H. F., III, Schreiner, P. R., Eds.; John Wiley & Sons:
Chichester, 1998; Vol. 2; pp 1070-1083.
(16) Chipot, C.; Pearlman, D. A. Mol. Sim. 2002, 28, 1-12.
(17) Knight, J. L.; Brooks, C. L., III J. Comput. Chem. 2009, 30,
1692-1700.
(18) Christ, C. D.; Mark, A. E.; van Gunsteren, W. F. J. Comput.
Chem. 2010, 31, 1569-1582.
(19) Kastner, J. WIREs Comput. Mol. Sci. 2011, 1, 932-942.
(20) Singh, S.; Chopra, M.; de Pablo, J. J. Annu. Rev. Chem. Biomol.
Eng. 2012, 3, 369-394.
(21) Beveridge, D. L.; DiCapua, F. M. Annu. Rev. Biophys. Biophys.
Chem. 1989, 18, 431-492.
(22) Kollman, P. A. Chem. Rev. 1993, 93, 2395-2417.
(23) Chipot, C. WIREs Comput. Mol. Sci. 2014, 4, 71-89.
(24) Straatsma, T. P.; Berendsen, H. J. C.; Postma, J. P. M. J. Chem.
Phys. 1986, 85, 6720-6727.
(25) MacKerell, A. D.; Wiorkiewicz-Kuczera, J.; Karplus, M. J. Am.
Chem. Soc. 1995, 117, 11946-11975.
(26) MacKerell, A. D.; et al. J. Phys. Chem. B 1998, 102, 3586-3616.
(27) Feller, S. E.; MacKerell, A. D. J. Phys. Chem. B 2000, 104, 7510-
7515.
(28) Hatcher, E.; Guvench, O.; MacKerell, A. D. J. Phys. Chem. B
2009, 113, 12466-12476.
(29) Brooks, B. R.; et al. J. Comput. Chem. 2009, 30, 1545-1614.
(30) Weiner, P. K.; Kollman, P. A. J. Comput. Chem. 1981, 2, 287-
303.
(31) Pearlman, D. A.; Case, D. A.; Caldwell, J. D.; Ross, W. S.;
Cheatham, T. E., III; DeBolt, S.; Fergusson, D.; Seibel, G.; Kollman, P. [ Pobierz całość w formacie PDF ]