摘要
蛋白质二级结构预测是后基因组学的重要内容,是能准确预测蛋白质分子三维空间结构的关键步骤,它被证明是个NP难问题。同时,蛋白质数据库中的数据是巨量的,每天以指数形式在增长,而且不同的数据库位于不同的地域,受不同组织管理。提出一个网格中基于结构分类和位矩阵编码的并行遗传算法,有效地解决这个NP难问题,并且充分利用网格强大的并行计算能力,提高预测效率和准确率。
Protein secondary structure prediction is an important content of post genome project and it is the key step of protein 3D structure prediction, and it is proved to be a NP-hard problem. The data in protein databases is a large amount, and it is increasing intensely everyday. Different protein databases locate in different places, and they are managed by different organizations. A parallel genetic algorithm is proposed based on structure classification and bit matrix coding to do the prediction, which can solve this NP-hard problem effectively, and it can make good use of the powerful parallel ability of grid to improve the efficient and accuracy of the prediction.
出处
《科学技术与工程》
2008年第5期1141-1145,共5页
Science Technology and Engineering
关键词
网格
蛋白质二级结构预测
结构分类
位矩阵编码
并行遗传算法
gird protein secondary structure prediction parallel genetic algorithm structure classification bit matrix coding
