IMPLEMENTATION OF THE ECEPP ALGORITHM, THE MONTE-CARLO MINIMIZATION METHOD, AND THE ELECTROSTATICALLY DRIVEN MONTE-CARLO METHOD ON THE KENDALL SQUARE RESEARCH KSR1 COMPUTER
Dr. Ripoll et al., IMPLEMENTATION OF THE ECEPP ALGORITHM, THE MONTE-CARLO MINIMIZATION METHOD, AND THE ELECTROSTATICALLY DRIVEN MONTE-CARLO METHOD ON THE KENDALL SQUARE RESEARCH KSR1 COMPUTER, Journal of computational chemistry, 16(9), 1995, pp. 1153-1163
In this article the adaptation of the Empirical Conformational Energy
Program for Peptides (ECEPP/3) and two conformational search methods [
viz., the Monte Carlo minimization (MCM) method and the electrostatica
lly driven Monte Carlo (EDMC) method] to the Kendall Square Research K
SR1 computer is described. The MCM and EDMC methods were developed to
surmount the multiple-minima problem in protein folding. Parallelizati
on of these codes led to substantial speedups (expressed as the ratio
between the mean time per energy evaluation in one processor and the m
ean time per energy evaluation in a set of processors) over the serial
versions of these codes. A comparison of the performance of these alg
orithms on the KSR1 and on the IBM ES9000 computers is presented. (C)
1995 by John Wiley & Sons, Inc.