The ff94 force field that is commonly associated with the AMBER simulation package is one of the most widely used parameter sets for biomolecular simulation. After a decade of extensive use and testing, limitations in this force field, such as over stabilization of α-helices, were reported by us and other researchers. This led to a number of attempts to improve these parameters, resulting in a variety of "AMBER" force fields and significant difficulty in determining which should be used for a particular application. We show that several of these continue to suffer from inadequate balance between different secondary structure elements. In addition, the approach used in most of these studies neglected to account for the existence in AMBER of two sets of backbone φ/ψ dihedral terms. This led to parameter sets that provide unreasonable conformational preferences for glycine. We report here an effort to improve the φ/ψ dihedral terms in the ff99 energy function. Dihedral term parameters are based on fitting the energies of multiple conformations of glycine and alanine tetrapeptides from high level ab-initio quantum mechanical calculations. The new parameters for backbone dihedrals replace those in the existing ff99 force field. This parameter set, which we denote ff99SB, achieves a better balance of secondary structure elements as judged by improved distribution of backbone dihedrals for glycine and alanine with respect to PDB survey data. It also accomplishes improved agreement with published experimental data for conformational preferences of short alanine peptides, and better accord with experimental NMR relaxation data of test protein systems.

We report unrestrained, all-atom molecular dynamics simulations of HIV-1 protease that sample large conformational changes of the active site flaps. In particular, the unliganded protease undergoes multiple conversions between the ''closed'' and ''semiopen'' forms observed in crystal structures of inhibitor-bound and unliganded protease, respectively, including reversal of flap ''handedness.'' Simulations in the presence of a cyclic urea inhibitor yield stable closed flaps. Furthermore, we observe several events in which the flaps of the unliganded protease open to a much greater degree than observed in crystal structures and subsequently return to the semiopen state. Our data strongly support the hypothesis that the unliganded protease predominantly populates the semiopen conformation, with closed and fully open structures being a minor component of the overall ensemble. The results also provide a model for the flap opening and closing that is considered to be essential to enzyme function.

Force-field validation is essential for the identification of weaknesses in current models and the development of more accurate models of biomolecules. NMR coupling and relaxation methods have been used to effectively diagnose the strengths and weaknesses of many existing force fields. Studies using the ff99SB force field have shown excellent agreement between experimental and calculated order parameters and residual dipolar calculations. However, recent studies have suggested that ff99SB demonstrates poor agreement with J-coupling constants for short polyalanines. We performed extensive replica-exchange molecular-dynamics simulations on Ala(3) and Ala(5) in TIP3P and TIP4P-Ew solvent models. Our results suggest that the performance of ff99SB is among the best of currently available models. In addition, scalar coupling constants derived from simulations in the TIP4P-Ew model show a slight improvement over those obtained using the TIP3P model. Despite the overall excellent agreement, the data suggest areas for possible improvement.

The use of parallel tempering or replica exchange molecular dynamics (REMD) simulations has facilitated the exploration of free energy landscapes for complex molecular systems, but application to large systems is hampered by the scaling of the number of required replicas with increasing system size. Use of continuum solvent models reduces system size and replica requirements, but these have been shown to provide poor results in many cases, including overstabilization of ion pairs and secondary structure bias. Hybrid explicit/continuum solvent models can overcome some of these problems through an explicit representation of water molecules in the first solvation shells, but these methods typically require restraints on the solvent molecules and show artifacts in water properties due to the solvation interface. We propose an REMD variant in which the simulations are performed with a fully explicit solvent, but the calculation of exchange probability is carried out using a hybrid model, with the solvation shells calculated on the fly during the fully solvated simulation. The resulting reduction in the perceived system size in the REMD exchange calculation provides a dramatic decrease in the computational cost of REMD, while maintaining a very good agreement with results obtained from the standard explicit solvent REMD. We applied several standard and hybrid REMD methods with different solvent models to alanine polymers of 1, 3, and 10 residues, obtaining ensembles that were essentially independent of the initial conformation, even with explicit solvation. Use of only a continuum model without a shell of explicit water provided poor results for Ala3 and Ala10, with a significant bias in favor of the α-helix. Likewise, using only the solvation shells and no continuum model resulted in ensembles that differed significantly from the standard explicit solvent data. Ensembles obtained from hybrid REMD are in very close agreement with explicit solvent data, predominantly populating polyproline II conformations. Inclusion of a second shell of explicit solvent was found to be unnecessary for these peptides.

The effects of the use of three generalized Born (GB) implicit solvent models on the thermodynamics of a simple polyalanine peptide are studied via comparing several hundred ns of well-converged replica exchange molecular dynamics (REMD) simulations using explicit TIP3P solvent to REMD simulations with the GB solvent models. It is found that when compared to REMD simulations using TIP3P, the GB REMD simulations contain significant differences in secondary structure populations; most notably an over-abundance of α-helical secondary structure. This discrepancy is explored via comparison of the differences in the electrostatic component of the free energy of solvation (ΔΔG Pol ) between TIP3P (via Thermodynamic Integration calculations), the GB models, and an implicit solvent model based on the Poisson Equation (PE). The electrostatic component of the solvation free energies are calculated using each solvent model for four representative conformations of Ala10. Since PE is found to have the best performance with respect to reproducing TIP3P ΔΔG Pol values, effective Born radii from the GB models are compared to effective Born radii calculated with PE (so-called perfect radii), and significant and numerous deviations in GB radii from perfect radii are found in all GB models. The effect of these deviations on the solvation free energy is discussed, and it is shown that even when perfect radii are used the agreement of GB with TIP3P ΔΔG Pol values does not improve. This suggests a limit to the optimization of the effective Born radius calculation, and that future efforts to improve the accuracy of GB must extend beyond such optimizations.

Parallel tempering or replica-exchange molecular dynamics (REMD) significantly increases the efficiency of conformational sampling for complex molecular systems. However, obtaining converged data with REMD remains challenging, especially for large systems with complex topologies. We propose a new variant to REMD where the replicas are also permitted to exchange with an ensemble of structures that have been generated in advance using high-temperature MD simulations, similar in spirit to J-walking methods. We tested this approach on two non-trivial model systems, a β-hairpin and a 3-stranded β-sheet and compared the results to those obtained from very long (>100 ns) standard REMD simulations. The resulting ensembles were indistinguishable, including relative populations of different conformations on the unfolded state. The use of the reservoir is shown to significantly reduce the time required for convergence.

We report unrestrained, all-atom molecular dynamics simulations of HIV-1 protease (HIV-PR) with a continuum solvent model that reproducibly sample closing of the active site flaps following manual placement of a cyclic urea inhibitor into the substrate binding site of the open protease. The open form was obtained from the unbound, semi-open HIV-PR crystal structure, which we recently reported (Hornak, V.; et al. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 915-920.) to have spontaneously opened during unrestrained dynamics. In those simulations, the transiently open flaps always returned to the semi-open form that is observed in all crystal structures of the free protease. Here, we show that manual docking of the inhibitor reproducibly induces spontaneous conversion to the closed form as seen in all inhibitor-bound HIV-PR crystal structures. These simulations reproduced not only the greater degree of flap closure, but also the striking difference in flap "handedness" between bound and free enzyme. In most of the simulations, the final structures were highly accurate. Root-mean-square deviations (RMSD) from the crystal structure of the complex were approximately 1.5 A (averaged over the last 100 ps) for the inhibitor and each flap despite initial RMSD of 2-5 A for the inhibitors and 6-11 A for the flaps. Key hydrogen bonds were formed between the flap tips and between flaps and inhibitor that match those seen in the crystal structure. The results demonstrate that all-atom simulations have the ability to significantly improve poorly docked ligand conformations and reproduce large-scale receptor conformational changes that occur upon binding.

The transferability of molecular mechanics parameters derived for small model systems to larger biopolymers such as proteins can be difficult to assess. Even for small peptides, molecular dynamics simulations are typically too short to sample structures significantly different than initial conformations, making comparison to experimental data questionable. We employed a PC cluster to generate large numbers of native and non-native conformations for peptides with experimentally measured structural data, one predominantly helical and the other forming a beta-hairpin. These atomic-detail sets do not suffer from slow convergence, and can be used to rapidly evaluate important force field properties. In this case a suspected bias toward alpha-helical conformations in the ff94 and ff99 force fields distributed with the AMBER package was verified. The sets provide critical feedback not only on force field transferability, but may also predict modifications for improvement. Such predictions were used to modify the ff99 parameter set, and the resulting force field was used to test stability and folding of model peptides. Structural behavior during molecular dynamics with the modified force field is found to be very similar to expectations, suggesting that these basis sets of conformations may themselves have significant transferability among force fields. We continue to improve and expand this data set and plan to make it publicly accessible. The calculations involved in this process are trivially parallel and can be performed using inexpensive personal computers with commodity components.