Accelerating the replica exchange method through an efficient all-pairs exchange

Abstract

The authors accelerate the replica exchange method through an efficient all-pairs replica exchange. A proof of detailed balance is shown along with an analytical estimate of the enhanced exchange efficiency. The new method provides asymptotically four fold speedup of conformation traversal for replica counts of 8 and larger with typical exchange rates. Experimental tests using the blocked alanine dipeptide demonstrate the method's correctness and show an approximate sampling efficiency improvement of 100% according to potential energy cumulative averages and an ergodic measure. An explicitly solvated PIN1 WW domain system of 4958 atoms is sampled using our new method, yielding a cluster sampling rate almost twice that of the single exchange near neighbor implementation. Computational software and scripts along with input and output data sets are available at.

Replica exchange statistical temperature Monte Carlo

“…Several sophisticated replica exchange methods [15][16][17][18][19][20][21][22][23][24][25][26][27] have been developed to resolve this limitation. We expect that the same strategy used in RESTMC represents one alternative route to overcome this limitation with a proper extension to molecular dynamics simulations.…”

“…The increased number of replicas requires more configurational swaps to sweep a whole temperature space and impairs the convergence. Several sophisticated REM variants have been developed [15][16][17][18][19][20][21][22][23][24][25][26][27] to resolve this system size dependence. Fukunishi et al 15 have developed the Hamiltonian REM, which performs replica exchanges between the original system and one with a scaled or deformed Hamiltonian.…”

Converting a High Performance Application to an Elastic Cloud Application

“…One way of overcoming energy barriers is to raise the temperature of the system; however, the paths obtained from high temperature simulations do not correspond to the paths at the lower temperature. A technique used to improve the sampling of the potential energy surface is parallel tempering, also called replica exchange molecular dynamics [7], [8], which has replicas at many temperatures. Many configurations are visited by the high temperature replicas and then annealed to lower temperature replicas by a Monte Carlo procedure that achieves the correct statistical distribution.…”

“…We also present a case study in converting a high performance application, a replica exchange molecular dynamics application [7], [8] implemented using MPI, to an elastic cloud application using a distributed computing framework that adheres to the presented guidelines. We compare the MPIbased high performance application with its corresponding elastic cloud application and show the challenges in running high performance computations, such as scalability, faulttolerance, failure recovery, are elegantly addressed with the elastic application.…”

Generalized Replica Exchange Method

“…[4][5][6][7][8] Simulating a set of replicas of the same system at a distribution of temperatures and swapping configurations among replicas offers an effective means of avoiding trapping in local minima and mitigating broken ergodicity at low temperatures. 9 A great deal of effort [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27] has been devoted to improving the efficiency of the standard tREM. One of the debated issues is its applicability to systems displaying strong phase transitions, around which metastable or unstable energy states intervene between two macroscopic phases.…”