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. General schemes for computing free energies of removing constraints: One of the largest prohibiting factors to accurate QM/MM free energies via the indirect approach is the configurational mismatch between the bonded and angle degrees of freedom between MM and QM/MM surfaces. As a result, it is highly desirable to incorporate bond and angle restraints on molecules of interest as to increase the overlap between high and low-level potential energy surfaces. Thus, we introduce a scheme that allows for the calculation of free energy cost associated with any number of bond and angle constraints, and capable of handling coupled degrees of freedom. Best Practices in QM-MM Modeling: Revisiting of the Double-Link Atom Approach Robust hybrid QM/MM methods are essential for characterizing and determining mechanisms of reactions inside of proteins. Describing the boundary between the quantum and classical regions requires a detailed consideration of the electrostatics in order to produce accurate energetics. Revisiting QM/MM methods we developed, we have implemented the double-link atom approach in the CHARMM to make the method readily available for users. Additionally, we have updated the CHARMM interface with open-source and commercial quantum chemistry software packages. Electrostatic parameters for amino acids in the Charmm36 forcefield have been reoptimized to balance the inadvertent polarization produced upon severing covalent bonds. We are currently applying this to investigate the several proposed mechanisms of Triose Phosphate Isomerase. Quantum Chemical Approaches for Predicting Physicochemical Properties Theoretical approaches can be a useful partner for predicting physiochemical properties important in experimental design, from drug discovery and development to studies in toxicology, as such predictive approaches can provide guidance in high-throughput screening for rational design. In quantum mechanical approaches, solvation can be accounted for explicitly or implicitly. We have designed several methods to predict properties in solution, including pKa and logP, that have been applied to two SAMPL blind challenges. For predicting the pKa, we have a scheme that uses the thermodynamic cycle with a higher-level energy correction capable of predicting pKa values within 0.6 pKa units from the experiment. Using the transfer free energies between water and octanol, we explored both standard density functional approximation methods and wavefunction methods that uses linear scaling algorithms that reduce the computational expense, achieving an MAD of 0.9 and 0.5 lopP units, respectively. Systematic fitting of Lennard-Jones interactions for representing polarization effects in additive force fields Additive force fields employ fixed point charges that are adjusted to a specific dielectric environment. The lack of electrostatic polarization hurts the accurate representation of water partitioning into a hydrophobic phase, e.g., a biological membrane, by 1-2.5 kcal/mol depending on the water model. We showed that an automated optimization of pairwise water-alkane parameters recovers experimental partition coefficients and helps to better describe passive water permeation through membranes. Development of generally applicable methods for calculating membrane permeabilities from molecular dynamics simulations Small-molecule permeability through biological membranes is an important parameter for drug efficacy and a natural target for molecular dynamics simulations. However, accurate calculations are hampered by deficiencies of additive force fields and challenged by slow relaxation of the lipid environment in response to enhanced sampling methods. In collaboration with the Laboratory of Membrane Biophysics, we studied and developed methods for permeability calculations. These methods include (a) counting of spontaneous permeation events, (b) maximum-likelihood estimation (MLE) of permeabilities via the inhomogeneous solubility-diffusion model, as well as (c) an extension of the MLE approach to trajectories coming from biased simulations. As a proof of concept, the methods were applied to water and oxygen permeation through different lipid bilayers. Results from methods (a)-(c) yielded consistent permeabilities, where the novel biasing approach (c) speeds up calculations by orders of magnitudes for both oxygen and water. The Homogeneity Condition: A Simple Way to Derive Isotropic Periodic Sum Potentials for Efficient Calculation of Long-Range Interactions in Molecular Simulation By using the isotropic periodic images of a local region to represent remote structures, long-range interactions become a function of the local conformation. This function is called the IPS potential, which folds long-ranged interactions into a short-ranged potential and can be calculated as efficiently as a cutoff method. Analytic solutions of IPS potentials have been solved for many interaction types. To further simplify the application of the IPS method, this work presents the homogeneity condition, which requires the sum of interaction energies for any particle is independent of cutoff distances for a truly homogeneous system. Using the homogeneity condition, one can avoid the complicated mathematical work to solve analytic solutions and can instead use simple functions as IPS potentials. Energies, volumes, and their fluctuations from these simulations demonstrate that simple IPS potentials obtained through the homogeneity condition can satisfactorily describe long-range interactions. A double exponential potential for van der Waals interaction Van der Waals (vdw) interaction is an important forces between atoms and molecules. Many potential functions have been proposed to model vdw interaction such as the Lennard-Jones (L-J) potential. To overcome certain drawbacks of existing function forms, this work proposes a double exponential (DE) potential that contains a repulsive exponential term and an attractive exponential term. This potential decays faster than the L-J potential and has a soft core. The DE potential is very flexible and its two exponential parameters can be adjusted continuously to mimic many potential functions. Accelerating Peptide Insertion in Membranes via Replica Exchange Molecular Dynamics with a Pulsed Electrical Field The interaction of proteins with cell lipid membranes is important for a wide variety of essential cellular processes, including signal transduction, membrane fusion, antimicrobial activity, etc. Molecular dynamics (MD) simulations allow the study of the structural and dynamic properties of this process with atomic-level detail. However, this process occurs on timescales which remain challenging to reach for MD simulations, even when employing advanced hardware such as GPUs. To improve sampling of peptide insertion events, we have employed replica exchange MD (REMD) that make use of a combination of increased temperature and application of an electric field. To overcome slow convergence of these REMD simulations due to too-favorable interactions of the peptide helical dipole with the field upon insertion (effectively preventing exchanges and slowing convergence), we have developed a pulsed field method where the strength of the electric field is smoothly and repeatedly increased to allow insertion and then decreased to improve convergence of the REMD simulations. This method will eventually be incorporated and freely available as part of the CHARMM software package. Development of Best-practices Articles for the Molecular Dynamics Community The goal of this workshop was to produce a series of best-practices articles.
Showing the most recent 10 out of 63 publications
