基于基因表达谱数据筛选差异表达基因新方法

发布时间：2018-05-01 13:15

本文选题：基因表达谱 + 差异表达基因　；参考：《数学的实践与认识》2016年18期

【摘要】：提出一种基于基因表达谱数据筛选差异表达基因的新方法;介绍了筛选差异表达基因常见方法-错误发现率方法(False Discovery Rate,FDR,),分析了多重假设检验p值性质,并根据p值性质提出了一种筛选差异表达基因新方法-单位γ度量法(Unit Measure-γ,UM-γ),建立了计算机模拟基因表达谱数据模型,制定了假阴性率、假阳性率、灵敏度、特异度以及总体错误率等作为考核指标,并使用基因表达谱模拟数据进行计算、比较;单位γ度量法估计非差异表达基因个数具有较高的稳定性和准确性;单位γ度量法既能够同时控制假阳性、假阴性以及总体错事件的发生,又能在一定程度上提高筛选结果的灵敏度和变异度;新提出的方法能有效、准确且稳定的对模拟数据差异表达基因进行筛选.
[Abstract]:A new method for screening differentially expressed genes based on gene expression data was proposed, and the common method of screening differential expression genes, False Discovery Rate (FDR), was introduced. The properties of P value of multiple hypothesis testing were analyzed, and a new method of screening differential expression genes, unit gamma measurement (Unit), based on the properties of P value (Unit), was proposed. Measure- gamma, UM- gamma), a computer simulated gene expression profile data model was established, and the false negative rate, false positive rate, sensitivity, specificity and total error rate were used as evaluation indexes, and the number of non differentially expressed genes was estimated by the unit gamma ray method. And accuracy; unit gamma measurement can not only control false positive, false negative and overall wrong events simultaneously, but also improve the sensitivity and variability of the screening results to a certain extent. The new method can be effective, accurate and stable for screening the differential expression basis of simulated data.

【作者单位】：南京医科大学附属南京医院医疗设备处;南京医科大学数学与计算机系;南京医科大学生物医学工程系;青岛理工大学土木工程系;
【基金】：南京医科大学科技发展基金面上项目(NJMU20150035)
【分类号】：Q811.4

【相似文献】