New theoretical techniques are being developed and characterized. These efforts are usually coupled with software development, and involve the systematic testing and evaluation of new ideas. This development is driven by current needs and interests. Multi-scale modeling methods and applications Important cellular events are too complex to be represented with high level quantum mechanical methods. Multi-scale modeling combines computationally expensive high level techniques such as quantum mechanical theory with relatively cheap low level ones such as molecular mechanics or coarse grained models. Usually multi-scale methods use high accuracy models to parameterize the coarser ones. However, it is also possible to couple these methods within a single framework to allow energies, forces, and NMA calculation to be done. With a general tool, the possible multiscale combinations are endless. MSCALE a generalized approach to multi-scale simulations. With the exception of QM/MM, multiscale methods typically are deployed sequentially with data from the high accuracy models being used to parameterize the coarser ones. However, coupling these disparate methods by generalizing the ideas underpinning additive and subtractive QM/MM schemes (Figure 4) is attractive. Within a single framework, support for both additive and subtractive approaches (or a combination of these) allows energy, force and normal mode calculations to be done with diverse sets of software and hardware. We have implemented and are currently validating this general approach in the CHARMM macromolecular simulation package thus supporting multiple, independent but connected calculations employing user-defined molecular subsystems. The list of functionality that is already supported includes: (1) Triple parallel scheme: (a) replicas, (b) hamiltonian term or layer distribution, and (c) parallelization of specific energy terms (2) Multi-scale Normal mode analysis via analytic Hessian calculations (3) Generalized approach to combining different model scales including the coupling of coarse-grained models, all-atom force fields, and quantum mechanical levels of theory. (4) General multiscale free energy perturbation between not only structural changes but also methodological ones. For example, computing the free energy differences between two different force fields or solvation models. (5) General method for combining different multiscale approaches (Figure 4) in a common framework. For example, combining both additive (via QM/MM) and subtractive (coarse-grained/all-atom) within a single calculation. This approach has been designed to run in a triple parallel scheme coupled with methods based on the REPLica functionality (the replica path method and its derivatives and replica exchange molecular dynamics). Exploring transition paths between two known structures A new method interpolates two elastic network potentials which are built from two known structures, and obtains a mixed potential with two minima and one saddle point. We analytically solved the saddle point of the mixed potential and did free-energy sampling by calculating the potential of mean-force (PMF). We explored various transition paths and applied this technique to explore the conformational transitions in kinesin, myosin, and chaperonin GroEL. Isotropic Periodic Sum method for long-range interaction calculation. The IPS method presents an accurate and efficient approach for calculating both electrostatic and van der Waals long-range interactions. It is demonstrated that the IPS method produces results very similar to that of Ewald summation, but with the following major advantages, (1) it eliminates unwanted symmetry artifacts raised from periodic boundary conditions, (2) it can be applied to potentials of any functional form and to fully and partially homogenous systems as well as finite systems, (3) it is more computationally efficient and can be easily parallelized for multiprocessor computers, (4) has been applied in L-J fluid and interface system studies. Recent developments and extensions greatly enhanced the accuracy and applicability of the IPS method. Osmolytes and denaturants: the Molecular Transfer Model Osmolytes and denaturants are widely used in experiments to study protein folding. Therefore understanding the mechanism by which they affect protein properties and predicting those effects is essential. Towards this end we have developed the Molecular Transfer Model (MTM). The MTM is the first molecular simulation technique capable of studying equilibrium folding and unfolding as a function of osmolyte type, osmolyte concentration, and pH. Previous computational studies used temperature to examine folding/unfolding. We have applied the MTM to (1) understand protein denatured state (DSE) properties under varying solution conditions, (2) determine the molecular origin of constant m-values, and (3) examine the accuracy of denatured state properties computed from single molecule FRET (Forster resonance energy transfer) data. We have also examined the mechanism of urea and guanidinium chloride action on hydrophobic and charge-charge interactions, forces that are important to protein stability, by computing the potential-of-mean force between two solutes at varying denaturant concentration. We found that these cosolutes denature proteins largely through direct hydrogen bonding between the cosolute molecules and the protein backbone. This was explicitly shown through simulations of the helix-coil transition in the presence of denaturants. Other ongoing method development projects: Exploring SCC-DFTB Paths for Mapping QM/MM Reaction Mechanisms Vibrational subsystem analysis: A method for probing free energies and correlations in the harmonic limit. Lipids: Improving the CHARMM Force Field and Experimental Collaborations. Ab Initio Modeling of Glycosyl Torsions and Anomeric Effects in a Model Carbohydrate: 2-Ethoxy Tetrahydropyran Folding simulations with explicit water Evaluation of Advanced Sampling Methodology Improved sampling using combined Conformation Space Annealing method with Replica Exchange Method Development of an effective sampling method for biomolecular systems based on the TIGER method.

Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
National Heart, Lung, and Blood Institute
Zip Code
Han, Kyungreem; Hudson, Phillip S; Jones, Michael R et al. (2018) Prediction of CB[8] host-guest binding free energies in SAMPL6 using the double-decoupling method. J Comput Aided Mol Des :
Zeng, Qiao; Jones, Michael R; Brooks, Bernard R (2018) Absolute and relative pKa predictions via a DFT approach applied to the SAMPL6 blind challenge. J Comput Aided Mol Des :
Damjanovic, Ana; Miller, Benjamin T; Okur, Asim et al. (2018) Reservoir pH replica exchange. J Chem Phys 149:072321
Leonard, Alison N; Simmonett, Andrew C; Pickard 4th, Frank C et al. (2018) Comparison of Additive and Polarizable Models with Explicit Treatment of Long-Range Lennard-Jones Interactions Using Alkane Simulations. J Chem Theory Comput 14:948-958
Narayan, Brajesh; Herbert, Colm; Yuan, Ye et al. (2018) Conformational analysis of replica exchange MD: Temperature-dependent Markov networks for FF amyloid peptides. J Chem Phys 149:072323
Wu, Xiongwu; Brooks, Bernard R (2018) Hydronium Ions Accompanying Buried Acidic Residues Lead to High Apparent Dielectric Constants in the Interior of Proteins. J Phys Chem B 122:6215-6223
Wang, Meiting; Li, Pengfei; Jia, Xiangyu et al. (2017) Efficient Strategy for the Calculation of Solvation Free Energies in Water and Chloroform at the Quantum Mechanical/Molecular Mechanical Level. J Chem Inf Model 57:2476-2489
Lee, Juyong; Konc, Janez; Janeži?, Dušanka et al. (2017) Global organization of a binding site network gives insight into evolution and structure-function relationships of proteins. Sci Rep 7:11652
Huang, Jing; Mei, Ye; König, Gerhard et al. (2017) An Estimation of Hybrid Quantum Mechanical Molecular Mechanical Polarization Energies for Small Molecules Using Polarizable Force-Field Approaches. J Chem Theory Comput 13:679-695
Li, Ying; Li, Hui; Pickard 4th, Frank C et al. (2017) Machine Learning Force Field Parameters from Ab Initio Data. J Chem Theory Comput 13:4492-4503

Showing the most recent 10 out of 63 publications