聚类GO术语在基因表达差异研究中的应用

Clustering GO Terms Applied to Differential Gene Expression Detection

针对基因功能分类体系基因本体(Gene Ontology,GO)特殊的有向无环图特点,改进传统的用单个GO术语检测基因差异表达信号的缺陷,设计出"聚类GO术语提升差异表达检测(ScaGO)"算法.通过简单的输入对照和实验组表达谱上的全部基因表达信号,来研究一些比较新的差异表达功能组,有助于进一步解释基因差异表达的生物学意义,如疾病发病机制、药物作用机理等.将ScaGO和基于单GO术语差异分析法应用到急性淋巴细胞性白血病数据集和酵母Rap1 DNA绑定突变体差异表达数据集上,结果显示,ScaGO能比基于单GO术语差异分析法发现一些新的与差异表达相关联的功能类基因,对于指导实验具有积极意义.

To improve individual GO term analysis algorithm for detecting differential gene expression,according to the directed acyclic graph structure property of gene classification system,Gene Ontology（GO）,a novel and effective method named significant cluster analysis based on GO（ScaGO） was presented.The inputs of ScaGO were the expression values from a case-control microarrary experiment,aimed at detecting some novel differential expression changes.The results had shown some insights into gene expression difference at the functional level,towarded clarification of the process of pathological changes or mechanism of medicine.Both ScaGO and individual GO term analysis were applied to the acute lymphoblastic leukemia expression dataset and yeast Rap1 DNA-binding mutant dataset.Compared to individual GO term analysis,ScaGO was turned out to be more sensitive,and some novel differential expression changes which were mostly reported were mined successfully.It means that our ScaGO can provide the positive help in the experimental guidance.Fig 1,Tab 3,Ref 21