论文部分内容阅读
通过比较不同位点致病性预测软件的预测效果和预测特点,从而根据错义突变所在疾病的类型和使用目的选择预测效果最佳的软件,为位点致病性预测软件的合理有效使用提供依据。将孟德尔疾病和复杂疾病作为数据集分类标准,使用经实验验证的致病错义突变数据集对7种位点致病性预测软件(Mutation Taster-2,Mutation Assessor,Poly Phen2,SIFT,LRT,Condel和Logit)的实际预测效果和特点进行比较分析。结果显示被测位点致病性预测软件的预测效用和特点各不相同,同时软件对于孟德尔疾病数据集的预测效果优于复杂疾病数据集;此外Condel和Logit软件作为组合型位点致病性预测软件,在孟德尔疾病和复杂疾病错义突变数据集中的预测效果均优于单一型位点致病性预测软件,但二者的敏感度和特异性差异也较大。在位点致病性预测软件的实际使用过程中,应根据错义突变所属疾病的类型和软件的特点选择最适软件,或同时使用多个预测特点互补的软件对位点的致病性进行综合预测。
By comparing the predictive and predictive characteristics of pathogenicity prediction software at different sites, the software with the best prediction effect is selected according to the type of the disease where the missense mutation occurs and the purpose of use, so as to provide a reasonable and effective use of the site pathogenicity prediction software in accordance with. Using Mendelian diseases and complex diseases as dataset classification criteria, seven sites of pathogenicity prediction software (Mutation Taster-2, Mutation Assessor, Poly Phen2, SIFT, LRT Condel and Logit) were compared and analyzed. The results showed that the predictive utility and characteristics of the pathogenicity prediction software were different, and the prediction results of the software for Mendelian disease data sets were superior to that of the complex disease data sets. In addition, Condel and Logit software were pathogenic as combination sites Predictors of predictive software were superior to single-site pathogenicity predictive software in Mendelian disease and complex disease missense mutation datasets, but their sensitivity and specificity were also different. During the actual use of the site-specific pathogenicity prediction software, the optimal software should be selected according to the type of the disease to which the missense mutation belongs and the characteristics of the software, or the pathogenicity of the site Comprehensive forecast.