项目名称: 多源基因表达数据横向整合的统计方法比较
项目编号: No.11526146
项目类型: 专项基金项目
立项/批准年度: 2016
项目学科: 数理科学和化学
项目作者: 王琳
作者单位: 首都经济贸易大学
项目金额: 3万元
中文摘要: 高通量基因组技术的发展带来了总量巨大、来源多样、特征多样的基因表达数据,在多源基因表达数据横向整合分析方面有很多不同的统计分析方法,如何针对不同特征的数据选择最佳的整合方法是一个重要的问题。本项目针对不同的数据特征,对不同的原始数据整合方法和结果整合方法从发现率、生物相关性、稳定性、稳健性四个方面进行比较,解决多源基因表达数据横向整合过程中如何针对不同特点的数据选择最为合适的数据整合方法的问题。本项目具体解决三个关键问题:(1)用于分类的数据特征的选择;(2)特定的数据特征的分类标准;(3)原始数据整合方法与结果整合方法之间的可比性问题。
中文关键词: 多源基因表达数据;差异共表达;数据整合;数据特征;
英文摘要: The rapid advances of various high-throughput technologies bring massive amount of gene expression data from different sources with different features. There are many horizontal integrative methods to analyze these gene expression data. How to choose the best integrative method according to different data features is a very important issue. This project compares the integrative methods including mega-analysis and meta-analysis according to different data features using four quantitative statistical evaluation criteria: detection capability, biological association, stability and robustness. This project aims to give the guideline to choose the most suitable method according to the data features in a given application. The major topics include: (1) the suitable data features for data classification; (2) the best criterion of classification for each data feature; (3) the comparability between mega-analysis methods and meta-analysis methods.
英文关键词: gene expression data from different sources;differential co-expression;integrative analysis;data features;