Lipschitzian optimization without the Lipschitz constant
Journal of Optimization Theory and Applications
Representations of quasi-Newton matrices and their use in limited memory methods
Mathematical Programming: Series A and B
Polypeptide structure prediction: real-value versus binary hybrid genetic algorithms
SAC '97 Proceedings of the 1997 ACM symposium on Applied computing
Implementing Genetic Algorithms with Sterical Constraints for Protein Structure Prediction
PPSN V Proceedings of the 5th International Conference on Parallel Problem Solving from Nature
The Journal of Machine Learning Research
Numerical Methods for Unconstrained Optimization and Nonlinear Equations (Classics in Applied Mathematics, 16)
Hi-index | 0.00 |
In this work we show an innovative approach to the protein folding problem based on an hybrid Immune Algorithm (IA) and a quasi-Newton method starting from a population of promising protein conformations created by the global optimizer DIRECT. The new method has been tested on Met-Enkephelin peptide, which is a paradigmatic example of multiple-minima problem, 1POLY, 1ROP and the three helix protein 1BDC. The experimental results show as the multistage approach is a competitive and effective search method in the conformational search space of real proteins, in terms of quality solution and computational cost comparing the results of the current state-of-art algorithms.