Protein Homology Modeling with Heuristic Search for Sequence Alignment

Molecular biology provides several important techniques to modern cancer research. Among them, protein structure prediction offers important insights towards the understanding of a protein's biochemical functions, which may eventually lead to the discovery of new cancer drugs. Homology modeling is an important technique for protein structure prediction. However, this modeling technique suffers from a misalignment between the template and the target sequences. The situation deteriorates when the sequence identity between the two is low. In this study, using the modeller program we applied three heuristic search algorithms, namely genetic algorithms, tabu search and particle swarm optimization, to align the template-target pair. Two model assessment scores, GA341 and DOPE, were used to guide the search process. A preliminary result showed that, under the same constraints on computational resources, genetic algorithms produced the best search result, and DOPE provided more effective assessment for the search