Search

Solving an N-body problem, electrostatic or gravitational, is a crucial task and the main computational bottleneck in manyscientific applications. Its direct solution is an ubiquitous showcase example for the compute power of graphics processingunits (GPUs). However, the naive pairwise summation hasOðN2Þcomputational complexity. The fast multipole method(FMM) can reduce runtime and complexity toOðNÞfor any specified precision. Here, we present a CUDA-accelerated,CþþFMM implementation for multi particle systems withr1potential that are found, e.g. in biomolecular simulations.The algorithm involves several operators to exchange information in an octree data structure. We focus on the Multipole-to-Local (M2L) operator, as its runtime is limiting for the overall performance. We propose, implement and benchmarkthree different M2L parallelization approaches. Approach (1) utilizes Unified Memory to minimize programming andporting efforts. It achieves decent speedups for only little implementation work. Approach (2) employs CUDA DynamicParallelism to significantly improve performance for high approximation accuracies. The presorted list-based approach(3) fits periodic boundary conditions particularly well. It exploits FMM operator symmetries to minimize both memoryaccess and the number of complex multiplications. The result is a compute-bound implementation, i.e. performance islimited by arithmetic operations rather than by memory accesses. The complete CUDA parallelized FMM is incorporatedwithin the GROMACS molecular dynamics package as an alternative Coulomb solver.

Receptor-ligand interactions are essential for biological function and their binding strength is commonly explained in terms of static lock-and-key models based on molecular complementarity. However, detailed information of the full unbinding pathway is commonly lacking due, in part, to the static nature of atomic structures and ensemble averaging inherent to bulk biophysics approaches. Here we combine molecular dynamics and high-speed force spectroscopy on the streptavidin-biotin complex to determine the binding strength and unbinding pathways over the widest dynamic range. Experiment and simulation show excellent agreement at overlapping velocities and provided evidence of the unbinding mechanisms. During unbinding, biotin crosses multiple energy barriers and visits various intermediate states far from the binding pocket while streptavidin undergoes transient induced fits, all varying with loading rate. This multistate process slows down the transition to the unbound state and favors rebinding, thus explaining the long lifetime of the complex. We provide an atomistic, dynamic picture of the unbinding process, suggesting that the lock-and-key mechanism may need revision in terms of many routes to the lock that might be relevant for other receptor-ligand bonds.

The dissociation of huperzine A (hupA) from Torpedo californica acetylcholinesterase (TcAChE) was investigated by 4-microsecond unbiased and biased all-atom molecular dynamics (MD) simulations in explicit solvent. We performed our study using memetic sampling (MS) for the determination of reaction pathways (RPs), metadynamics to calculate free energy, and maximum-likelihood estimation (MLE) to recover kinetic rates from unbiased MD simulations. Our simulations suggest that the dissociation of hupA occurs mainly via two RPs: a front-door along the axis of the active-site gorge (pwf) and through a new transient side-door (pws), i.e., formed by the Omega-loop (residues 67--94 of TcAChE). Analysis of the inhibitor unbinding along the RPs suggests that pws is opened transiently after hupA and the Omega-loop reach a low free-energy transition state characterized by the orientation of the pyridone group of the inhibitor directed toward the Omega-loop plane. Unlike pws, pwf does not require large structural changes of TcAChE to be accessible. The estimated free energies and rates agree well with available experimental data. The dissociation rates along the unbinding pathways are similar, suggesting that the dissociation of hupA along pws is likely to be relevant. This indicates that perturbations to hupA-TcAChE interactions could potentially induce pathway hopping. In summary, out results characterize the slow-onset inhibition of TcAChE by hupA, which may provide the structural and energetic basis for the rational design of the next-generation slow-onset inhibitors with optimized pharmacokinetic properties for the treatment of Alzheimer's disease.

Phage shock protein A (PspA) belongs to the highy conserved PspA/IM30 family and is a key component of the stress inducible Psp system in Escherichia coli. One of its central roles is the regulatory interaction with the transcriptional activator of this system, the σ54 enhancer binding protein PspF, a member of the AAA+ protein family. The PspA/F regulatory system has been intensively studied and serves as a paradigm for AAA+ enzyme regulation by trans-acting factors. However, the molecular mechanism of how exactly PspA controls the activity of PspF and hence σ54-dependent expression of the psp genes is still unclear. To approach this question, we identified the minimal PspF-interacting domain of PspA, solved its structure, determined its affinity to PspF and the dissociation kinetics, identified residues that are potentially important for PspF regulation and analyzed effects of their mutation on PspF in vivo and in vitro. Our data indicate that several characteristics of AAA+ regulation in the PspA·F complex resemble those of the AAA+ unfoldase ClpB, with both proteins being regulated by a structurally highly conserved coiled-coil domain. The convergent evolution of both regulatory domains points to a general mechanism to control AAA+ activity for divergent physiological tasks via coiled-coil domains.

Förster Resonance Energy Transfer (FRET) experiments probe molecular distances via distance dependent energy transfer from an excited donor dye to an acceptor dye. Single molecule experiments not only probe average distances, but also distance distributions or even fluctuations, and thus provide a powerful tool to study biomolecular structure and dynamics. However, the measured energy transfer efficiency depends not only on the distance between the dyes, but also on their mutual orientation, which is typically inaccessible to experiments. Thus, assumptions on the orientation distributions and averages are usually made, limiting the accuracy of the distance distributions extracted from FRET experiments. Here, we demonstrate that by combining single molecule FRET experiments with the mutual dye orientation statistics obtained from Molecular Dynamics (MD) simulations, improved estimates of distances and distributions are obtained. From the simulated time-dependent mutual orientations, FRET efficiencies are calculated and the full statistics of individual photon absorption, energy transfer, and photon emission events is obtained from subsequent Monte Carlo (MC) simulations of the FRET kinetics. All recorded emission events are collected to bursts from which efficiency distributions are calculated in close resemblance to the actual FRET experiment, taking shot noise fully into account. Using polyproline chains with attached Alexa 488 and Alexa 594 dyes as a test system, we demonstrate the feasibility of this approach by direct comparison to experimental data. We identified cis-isomers and different static local environments as sources of the experimentally observed heterogeneity. Reconstructions of distance distributions from experimental data at different levels of theory demonstrate how the respective underlying assumptions and approximations affect the obtained accuracy. Our results show that dye fluctuations obtained from MD simulations, combined with MC single photon kinetics, provide a versatile tool to improve the accuracy of distance distributions that can be extracted from measured single molecule FRET efficiencies.

Here we report on the performance of GROMACS 4.6 on the SuperMUC cluster at the Leibniz Rechenzentrum in Garching. We carried out benchmarks with three biomolecular systems consisting of eighty thousand to twelve million atoms in a strong scaling test each. The twelve million atom simulation system reached a performance of 49 nanoseconds per day on 32,768 cores.

Within molecular dynamics simulations of protein–solvent systems the exact evaluation of long-range Coulomb interactions is computationally demanding and becomes prohibitive for large systems. Conventional truncation methods circumvent that computational problem, but are hampered by serious artifacts concerning structure and dynamics of the simulated systems. To avoid these artifacts we have developed an efficient and yet sufficiently accurate approximation scheme which combines the structure-adapted multipole method (SAMM) [C. Niedermeier and P. Tavan, J. Chem. Phys., 101, 734 (1994)] with a multiple-time-step method. The computational effort for MD simulations required within our fast multiple-time-step structure-adapted multipole method (FAMUSAMM) scales linearly with the number of particles. For a system with 36,000 atoms we achieve a computational speed-up by a factor of 60 as compared with the exact evaluation of the Coulomb forces. Extended test simulations show that the applied approximations do not seriously affect structural or dynamical properties of the simulated systems.

Proteins are complex macromolecules that evolved over billions of years to be active in aqueous solution. Water is a key element that stabilizes their structure, and most structural studies on proteins have thus been carried out in aqueous environment. However, recent experimental approaches have opened the possibility to gain structural information on proteins from gas-phase measurements. The obtained results revealed significant structural memory in proteins when transferred from water to the gas phase. However, after several years of experimental and theoretical research, the nature of the structural changes induced by vaporization, the exact characteristics of proteins in the gas phase, and the physicochemical forces stabilizing dehydrated proteins are still unclear. We will review here these issues using both experimental and theoretical sources of information

Enoyl-CoA carboxylases/reductases (ECRs) are some of the most efficient CO2-fixing enzymes described to date. However, the molecular mechanisms underlying the extraordinary catalytic activity of ECRs on the level of the protein assembly remain elusive. Here we used a combination of ambient-temperature X-ray free electron laser (XFEL) and cryogenic synchrotron experiments to study the structural organization of the ECR from Kitasatospora setae. The K. setae ECR is a homotetramer that differentiates into a pair of dimers of open- and dosed-form subunits in the catalytically active state. Using molecular dynamics simulations and structure-based mutagenesis, we show that catalysis is synchronized in the K. setae ECR across the pair of dimers. This conformational coupling of catalytic domains is conferred by individual amino acids to achieve high CO2-fixation rates. Our results provide unprecedented insights into the dynamic organization and synchronized inter- and intrasubunit communications of this remarkably efficient CO2-fixing enzyme during catalysis., NSF Science and Technology Center Grant; Biology with X-ray Lasers, BioXFEL; Scientific and Technological Research Council of Turkey (TÜBİTAK); 2232 International Fellowship for Outstanding Researchers Program; 1001 Scientific and Technological Research Projects Funding Program; U.S. Department of Energy; Office of Science; Office of Basic Energy Sciences; European Union (EU); Horizon 2020 ; European Research Council (ERC); SYBORG; U.S. Department of Energy Joint Genome Institute; DOE Office of Science User Facility; Max Planck Society; a Max-Planck-Partner Group; CONICYT; CONICYT Doctorado Nacional Grant; Fondo Nacional de Desarrollo Científico y Tecnológico (Fondecyt) Postdoctoral Fellowship; DOE Office of Science, Biological Environmental Research; National Institutes of Health; National Science Foundation Major Research Instrument Grant; NIGMS; Stanford PRECOURT Institute