项目名称: 癌症生物标记识别的基因网络研究
项目编号: No.61272315
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 陆慧娟
作者单位: 中国计量学院
项目金额: 80万元
中文摘要: 在后基因组时代,高通量生物数据的积累使得系统分析与癌症相关内在调控网络成为可能,这些分析对发现用于癌症诊断和治疗的关键分子标记具有重要意义。本项目以癌症生物标记识别为研究背景,将主要通过计算方法来提高基因网络预测的可靠性,使之满足异构、高维、小样本的基因组数据分析的需要。我们将以蛋白质和基因表达数据,蛋白质互作数据为基础,通过多源数据融合、基因网络模型重建、子网分析等机器学习方法来识别癌症发生相关重要基因及功能模块。该研究课题将开发适合复杂大规模基因网络构建和分析的算法,建立相关生物信息学平台。结合公开的真实生物学数据,发现与特定癌症相关的分子标记及其基因调控途径,采用临床数据做进一步验证。 本项目的研究将提供肿瘤生物学中分子标记的预测和识别的算法及应用平台。从生物信息学角度探索癌症早期诊断和个性化分析的方法,为分子生物学的发展提供理论和技术支持。
中文关键词: 基因调控网络;基因表达数据;机器学习;数据挖掘;癌症诊断
英文摘要: In post-genomics era, with the accumulations of high-throughput biological data, it became possible to systematically analyze the gene regulatory networks inherent in cancer; and this process is essential for identification of molecular biomarkers which will be used to cancer prognosis and treatment. Most of proteomics or genomics data contain large numbers of genes but with very limited sample sizes. This poses big challenges for statistical models and algorithms. To address this problem, we will investigate computational approaches to integrate heterogeneous "omics" data and improve the reliability of gene network prediction algorithms. In this project, we will utilize and integrate transcriptome, proteome and protein interaction data, combine multiple gene network models with sub-network analysis to identify the related genes and functional modules in cancer. We will develop novel algorithms that will be suituble for large-scale complex gene network reconstruction and analysis, and set up a public accessible bioinformatics web site for tumor biology research. Based on this platform, we will use real biological data to discover potential molecular biomarkers and pathways for specific cancers, and further validate those results using clinical data. The study will provide new algorithms and a web site for tumor
英文关键词: Genes Regulatory Network;Gene Expression Data;Machine Learning;Data Mining;Cancer Diagnosis