Comparing search strategies for finding global optima on energy landscapes
✍ Scribed by Foreman, K. W.; Phillips, A. T.; Rosen, J. B.; Dill, K. A.
- Publisher
- John Wiley and Sons
- Year
- 1999
- Tongue
- English
- Weight
- 150 KB
- Volume
- 20
- Category
- Article
- ISSN
- 0192-8651
No coin nor oath required. For personal study only.
✦ Synopsis
We provide some tests of the convex global underestimator Ž . CGU algorithm, which aims to find global minima on funnel-shaped energy landscapes. We use two different potential functions-the reduced Lennard᎐Jones cluster potential, and the modified Sun protein folding potential, to compare the CGU algorithm with the simplest versions of the traditional Ž . trajectory-based search methods, simulated annealing SA , and Monte Carlo Ž . MC . For both potentials, the CGU reaches energies lower on the landscapes than both SA and MC, even when SA and MC are given the same number of starting points as in a full CGU run or when all methods are given the same amount of computer time. The CGU consistently finds the global minima of the Lennard᎐Jones potential for all cases with up to at least n s 30 degrees of freedom. Finding the global or near-global minimum in the CGU method w Ž 3 . Ž 4 .x requires polynomial time scaling between O n and O n , on average.