Abstract: As one of the severe diseases, HBV（Hepatitis B Virus） infection is seriously affecting human health. This kind of virus infection is the main reason that leads to chronic liver disease, cirrhosis and liver cancer. Due to the particularity of HBV replication and high variability characteristics, related studies have revealed that the HBV gene mutation is the basic reason of persistent HBV infection. In order to understand the genetic variation of HBV, the SNP detection from HBV sequences has been widely applied in the large number of research, the detected SNP loci may contain great clinical significance. However, currently, the SNP loci detection methods are restricted by some negative factors, such as high technical difficulty, high expense and so on. Therefore, to explore a computer-based method for SNP loci detection becomes a trend. In this paper, considering the characteristics of SNP loci of the HBV sequence, an method of SNP loci detection based on optimal risk and prevention pattern is proposed. The proposed method is first applied to detect the SNP site in the HBV sequence. Experimental results show that the method has not only effectively detected the SNP loci of the sequence on HBV X gene fragment and the pre-C gene fragment which have been reported, and has also found a new SNP loci. Different from the SNP loci detection with hardware, the proposed method has the advantages of simple operation, low cost, and it can be accepted by general laboratory and medical institutions.

Key words: Hepatitis B Virus, feature selection, optimal risk and preventive patterns, Single Nucleotide Polymophism（SNP）

摘要： 乙型肝炎病毒（Hepatitis B Virus，HBV）感染作为严重影响人类健康的疾病之一，是导致慢性肝脏疾病、肝硬化和肝癌的主要元凶。HBV由于其自身复制的特殊性，具有高变异特性，据研究表明HBV基因变异是HBV持续感染的根本原因。为了了解HBV的基因变异情况，检测HBV序列的SNP位点即单突变位点已广泛应用于大量的研究，所检测出的SNP位点对指导临床有重要意义。但是目前关于SNP位点检测的方法多因技术难度较高，费用大等不利因素而受到制约。因此，探讨一种基于计算机的SNP位点检测方法成为一种趋势。针对HBV序列的 SNP位点的特点，提出了一种基于最优风险与预防模型的HBV序列的SNP位点检测方法。方法首次应用于HBV序列的SNP位点检测，实验结果表明：该方法不仅有效地检测出HBV序列的X基因片段和前C区基因片段中已经报道的位点，而且还发现了一些新的SNP位点。与硬件检测SNP位点不同的是，所提出的计算机方法具有操作简单和费用低的优点，而且普通实验室和医疗机构均可以承受。

关键词: 乙型肝炎, 特征选择, 风险与预防模式, 单核苷酸多态性（SNP）