摘要
蛋白质折叠问题就是从氨基酸序列中预测蛋白质的构象,该问题是生物信息学的一个突出问题。主要研究二维HP格点模型,它是用于模拟蛋白质折叠问题的一个具有代表性的简化模型,并且将蚁群算法用于求解该二维HP蛋白质的折叠问题。此外,在局部搜索机制中引入一种改进的牵引移动方法,这是一个提高蛋白质构象的有效方法。实验结果表明,针对较长的氨基酸序列,改进的带牵引移动的蚁群算法(ACO+)比ACO能够获得更低能量的构象,证明了所提出的改进蚁群算法是预测蛋白质结构的有效方法。
Protein folding problem is to predict a protein＇s conformation from its amino acid sequence,it is one of the most prominent problems in bioinformatics. In this paper we mainly study two-dimensional（ 2D） hydrophobic-polar（ HP） lattice model,which is a representative simplified model widely used for simulating the protein folding problem,and apply the ant colony optimisation（ ACO） to solving the 2D HP protein folding problem. Besides,we introduce an improved pull moves method to local search mechanism,to improve protein＇s conformation it is an effective method. Experimental results show that the improved ACO with pull moves（ ACO ＋） can obtain lower-energy conformations than ACO for several longer amino acid sequences,this demonstrates that our ACO ＋ algorithm is a effective method to predict the structure of protein.
出处
《计算机应用与软件》
CSCD
2015年第3期275-278,共4页
Computer Applications and Software
基金
甘肃省教育厅研究生导师项目(1215_04)
