摘要
针对单序列蛋白质二级结构预测问题,提出了一种基于高斯先验最大熵(GP-MaxEnt)模型的预测方法.该方法根据氨基酸的构象偏好进行特征构造,利用改进迭代缩放算法(ⅡS)训练高斯先验最大熵模型.使用CB513数据集对GP-MaxEnt模型进行了测试分析.试验表明,该方法简单有效,能够获得较好的预测精度.
Aimed at solving the problem of single-sequence protein secondary structure prediction, a novel method based on Gaussian prior maximum entropy （GP-MaxEnt） model is proposed. In this method, the feature construction was firstly performed based on the conformational preference of amino acid residues, and the improved iterative scaling （IIS） method was used to train the GP-MaxEnt model. CB513 dataset was employed to test this model. The experimental results indicate that the proposed method is effective and can achieve better results in predictive accuracy.
出处
《哈尔滨工业大学学报》
EI
CAS
CSCD
北大核心
2009年第8期65-68,共4页
Journal of Harbin Institute of Technology
基金
国家自然科学基金资助项目(60571025)
国家高技术研究发展计划资助项目(2006AA01Z308)
关键词
蛋白质二级结构预测
单序列预测
最大熵模型
高斯先验
protein secondary structure prediction
single-sequence prediction
maximum entropy model
Gaussian prior
