A differential evolution approach for protein structure optimisation using a 2D off-lattice model

摘要

Protein structure optimisation is a well-known problem in bioinformatics. This work applies an evolutionary algorithm to solve the protein structure optimisation problem based on the AB off-lattice model. Three different implementations of the differential evolution (DE) algorithm were developed, a sequential and two parallel. The parallel implementations (master-slave and ring-island) showed superior performance than the sequential one. Experiments were done using a benchmark of toy sequences with 13 to 55 monomers long. Results of the DE implementations were compared with other works in the literature. Good results were achieved for most sequences, not achieving the optimal values, but competitive with other specialised methods. Overall results encourage further research towards the use of knowledge-based operators to improve performance of DE.