In conjunction with our development of new methods we are actively engaged in developing or contributing to the development of several simulation and modeling software packages. We make our development efforts freely available so as to maximize the impact of our work. Improved Replica Exchange Molecular Dynamics Functionality in CHARMM While replica exchange molecular dynamics (REMD) functionality has been present in CHARMM (via the REPD module) for several versions, its usefulness was limited for the following reasons: 1) it could not be used with the faster DOMDEC code, 2) only one parallel thread per replica could be used, severely limiting its scalability, and 3) exchanges in the NPT ensemble were not handled correctly. We have overhauled the REPD module of CHARMM to address all of these limitations, allowing large-scale REMD simulations of biological systems, including those requiring NPT (e.g. membranes). In addition, we found and fixed an error in the exchange criterion calculation for temperature REMD in 2D that was preventing it from working correctly. This code is currently undergoing rigorous beta testing and will be submitted for the next major release of CHARMM. Extension of CPPTRAJ Analysis Software for CHARMM While CPPTRAJ is primarily distributed with the Amber biomolecular software package, it has for several years had the ability to process CHARMM trajectories as well. Despite this, there are several powerful analysis capabilities that CPPTRAJ has that were not available to CHARMM users, namely the ability to process ensembles of trajectories (from e.g. REMD simulations). CPPTRAJ has now been extended to process ensembles of trajectories from CHARMM and read CHARMM REMD logs for advanced analysis of REMD performance. In addition, CPPTRAJ can now read more CHARMM XYZ file formats (including CHARMM coordinates and restart formats), and extract energies from CHARMM output, increasing its utility to the wider MD community. Contributions to the development of Amber Molecular simulation and modeling software packages are the vehicle for computational research and experiment. Implementation of new methods and options is the key to facilitate cutting edge researches. In recent years, this lab has developed a series new compuatational methods, such as the self-guided Langevin dynamics for efficient conformational searching and sampling, the isotropic periodic sum method for accurate and efficient calculation of long-range interactions, and the map-based modeling tool, EMAP, for electron microscropy studies. Implementation of these new methods enables researchers to tackle difficult problems. These methods have been implemented into CHARMM to expand its capability in molecular simulation, conformational search, and structure prediction. These methods are all available in CHARMM version 41. These methods are also been implemented into another widely used simulation package, AMBER, to extend the user scope to access these methods. The SGLD, IPS, and EMAP methods are available in AMBER version 16. Dispersion PME The use of cutoffs to truncate dispersion interactions is problematic in highly anisotropic systems. To overcome this problem, we have implemented a particle mesh Ewald scheme for dispersion in the OpenMM and CHARMM packages, including the highly parallel DOMDEC routines. These developments allow for development and testing of the next generation CHARMM lipid force field, and ensure that they will be widely supported in software packages when published. Crucially, our implementation allows for the existing combination and exclusion rules to be used, so the long-range behavior can be introduced to the existing force field as a minimal perturbation. Dispersion PME will be a standard for future CHARMM force fields. Contributions to the development of Psi4 and OpenMM Many new developments towards multi-scale modeling have been advanced through our work in the Psi4 QM code as well as the OpenMM MM code. By redesigning the Psi4 code to be driven as a Python module, we have increased its interoperability and are working to incorporate it into our QM/MM workflow. The OpenMM package uses a similarly modern design and our developments of dispersion PME, fast polarization methods and new integrators have expanded the range of tools available to the community. Both packages have recently published update papers, and are both freely available under thehighly permissive LGPL license. In terms of hardware and software, the LOBOS system has acquired the following over the past year: two NVIDIA P100 GPU based systems, four NVIDIA Consumer grade Titan Xp system for development and code testing,and an Intel Knights Landing system for Andy/Samars porting/benchmarking work. On the administrative software side, conversion to the SLURM queueing system has resulted in better system utilization and more efficient use of hardware (better cpu and GPU affinity support) and. SLURM is used by many major computing sites so this will ease the learning curve for users coming from other sites and enhance portability of collaboratorspipelines to and from our system.
Showing the most recent 10 out of 15 publications
