中文摘要：

在后基因组时代，单核苷酸多态性（Single Nucleotide Polymorphism，SNP）数据的快速积累使得用计算方法对编码区和非编码区SNP进行功能分析成为可能；但是目前各种SNP功能预测方法的预测精度均不能令人满意。针对这一问题，文章根据SNP的分类，分别对错义SNP、同义SNP和非编码区SNP功能分析的生物信息学方法进行了总结。其中重点介绍了基于序列特征和基于结构特征的有害错义SNP预测方法。还介绍了互联网上提供SNP资源和功能注释工具。分析表明，各种SNP功能分析方法还有很大的改进余地。

英文摘要：

In the post - genomic area, the fast accumulation of SNP data available in the public domain in recent years enabled the in silico functional analysis of the SNPs which are located both in coding and non - coding regions. However, the accuracy of current SNP functional prediction tools is still not satisfactory. According to the types of SNPs, a broad review of Bioinformatics methods available is presented to understand the functional effects of genetic variants on the gene products. Special attention has been drawn on the deleterious non- synonymous SNPs prediction based on sequence features and structure features. Also, public SNP resources and SNP functional annotation tools are summarized. The analysis demonstrates that there are a lot of unsolved problems and innovations in this field.