(685e) Ssages: A Comprehensive Platform for Enhanced Sampling Simulations | AIChE

(685e) Ssages: A Comprehensive Platform for Enhanced Sampling Simulations


Whitmer, J. - Presenter, University of Notre Dame
Colón, Y. J., Argonne National Laboratory
Sidky, H., University of Notre Dame
de Pablo, J. J., University of Chicago
Sikora, B. J., Northwestern University
Helfferich, J., Karlsruhe Institute of Technology
Sevgen, E., University of Chicago
Moller, J., University of Chicago
Guo, A., University of Chicago
Bezik, C., University of Chicago
Lequieu, J., University of Chicago
Webb, M., University of Chicago
Ramezani-Dakhel, H., University of Chicago
Rahimi, M., University of Chicago
Thapar, V., Chonnam National University, South Korea
Li, J., University of Chicago
Reid, D., University of Chicago
Quevillon, M., University of Notre Dame
Gygi, F., University of California--Davis
Galli, G., University of Chicago
Heinonen, O. G., Argonne National Laboratory
Ferrier, N. J., University of Wisconsin-Madison
Jiang, X., The University of Chicago
Molecular simulations are an essential tool in modern research, guiding experimental efforts and providing fundamental insights into underlying phenomena. Well-designed simulations offer a molecule-scale microscope with exquisite control over the physical phenomena of interest. While immensely powerful, the simulation toolbox has some important limitations—the accuracy of computations requires proper sampling of the system. While this may be achieved by simulating long trajectories or multiple independent copies of a system, the size of the systems and the time-scales that are accessible are still restricted by hardware. Importantly, systems that have rough energetic landscapes can become “stuck”, impeding proper exploration of configurations.

A palette of techniques has evolved to overcome these difficulties through static and dynamic modifications of the statistical ensemble underlying the sampling; these may be restriction based (as in umbrella sampling and replica exchange techniques) or on-the-fly adaptations (as in the powerful adaptive biasing force and metadynamics methods). Regrettably, the implementation of these techniques is not always straightforward and can be unattainable to users. This has often constrained newly developed, ostensibly general methods to problem or software specific applications, preventing the use of these techniques across platforms and research.

With these concerns in mind, we have developed SSAGES: Software Suite for Advanced General Ensemble Simulations. SSAGES is a multi-platform software that allows the use of enhanced sampling techniques on molecular simulations, which allows users to seamlessly apply enhanced sampling techniques to a variety of systems using widely used simulation engines such as LAMMPS, GROMACS, QBox, and OpenMD. We leverage SSAGES to study a variety of systems using multiple methods and discuss relative merits of flat histogram and string methods. In this talk, we will demonstrate the powerful methods, such as metadynamics, string methods, forward flux sampling and adaptive biasing force to several example systems, including dissociation of NaCl, isomerization of alanine dipeptide, mechanical properties of liquid crystals, self-assembly of porous materials, and discuss applications to related classes of problems.