合成金字塔预测模型中内含的改进型CBA预测方法  

Improved CBA prediction algorithm in compound pyramid model

作　　者：杨炳儒[1] 周谆[1] 侯伟[1] 

机构地区：[1]北京科技大学信息工程学院,北京100083

出　　处：《计算机应用研究》2009年第12期4617-4620,共4页Application Research of Computers

基　　金：国家自然科学基金重点资助项目(69835001);国家教育部科技重点资助项目([2000]175);北京市自然科学基金资助项目(4022008)

摘　　要：蛋白质二级结构预测问题,是生物信息学领域中最为重要的任务之一,历经三十多年的研究,已取得了一些进展,尤其是近来集成预测模型与混合预测模型的引入,为预测精度带来了一定程度的提高,然而其离从二级结构推导三级结构的目标,仍然存在很大差距。为了有效提高蛋白质二级结构预测精度,以KDTICM理论的扩展性研究与KDD*模型为基础,使用基于KDD*模型的关联分析蛋白质二级结构预测方法KAAPRO,提出一种基于支持度与可信度的复杂距离度量的CBA(classification based on association)算法,并以该算法为核心构建逐步求精、多层递阶的合成金字塔模型,该模型整体贯穿领域知识,并采用因果细胞自动机选择有效物化属性。在对偏alpha、beta型蛋白质的预测实验中,改进型CBA算法较好地完成了对结构特征不明显氨基酸的预测,获得了较优的预测效果。The problem of protein secondary structure prediction is one of the most important problems in bioinformatics. After the study of this problem for 30 years and more, there have been some breakthroughs. Especially the introduction of ensemble prediction model and hyrid prediction model, make the accuracy of prediction better, but there is a long distance to induce the tertiary structures from the secondary ones. As one of the researches of KDTICM theory, this paper proposed an improved algorithm of CBA , which was based on KDD ＊ model and combined with KAAPRO method, for protein secondary structure prediction. And proposed a gradually enhanced, multi-layer systematic perditions model, compound pyramid mode. The kernel of this model was association rules analysis of CBA. Domain knowledge was used through the whole model, and the phychemical attributes was chosen by causal cellular automata. The experiment predicted the proteins containing more alpha/beta structure. The structures of amino acids, whose structural traits were obscure, were predicted well by the improved CBA . Hence, the result of this model is satisfying too.

关 键 词：关联规则 蛋白质二级结构预测 KDD＊ 合成金字塔模型 基于关联分类算法 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]