结合基因本体论和计算机模式识别技术筛选应答神经损伤的重要效应基因

Key Effect Genes Responding to Nerve Injury Identified by Gene Ontology and Computer Pattern Recognition

目的利用基因本体论(gene ontology,GO)方法结合计算机模式识别技术从基因芯片的海量数据中筛选出应答神经损伤的重要基因,并选择一种可能的关键基因Cd44对生物信息学结果进行验证。方法利用MATLAB软件对基因芯片数据(GSE26350)进行数据挖掘和GO分析,利用不同功能模块的基因在主成分得分图上的位置分布,从得出的核心基因分子谱中选择Cd44。干扰Cd44与其配体硫酸软骨素C(chondroitin sulfate C,CSC)的正常结合,观察其对神经突起延伸的影响。结果 GO分析结果显示,主成分得分图上红色*标记的第1类基因在分子功能方面主要与信号传递、受体活动和蛋白结合等相关。Cd44是此类基因6个外围效应蛋白基因之一,其功能多样。向培养的神经元培养基中添加不同试剂干扰CSC与Cd44的正常结合可不同程度缓解CSC对神经突起延伸的抑制作用。结论本实验初步证明CSC与筛选出的神经元表面分子Cd44的结合可抑制神经突起的延伸,这说明利用GO分析基因芯片可筛选出特定生理过程中的重要分子。

Objective In order to screen out important genes from large gene data of gene microarray after nerve injury,we combine gene ontology（GO） method and computer pattern recognition technology to find key genes responding to nerve injury,and then verify one of these screened-out genes.Methods Data mining and gene ontology analysis of gene chip data GSE26350 was carried out through MATLAB software.Cd44 was selected from screened-out key gene molecular spectrum by comparing genes＇ different GO terms and positions on score map of principal component.Function interferences were employed to influence the normal binding of Cd44 and one of its ligands,chondroitin sulfate C（CSC）,to observe neurite extension.Results Gene ontology analysis showed that the first genes on score map（marked by red ＊） mainly distributed in molecular transducer activity,receptor activity,protein binding et al molecular function GO terms.Cd44 is one of six effector protein genes,and attracted us with its function diversity.After adding different reagents into the medium to interfere the normal binding of CSC and Cd44,varying-degree remissions of CSC＇s inhibition on neurite extension were observed.Conclusion CSC can inhibit neurite extension through binding Cd44 on the neuron membrane.This verifies that important genes in given physiological processes can be identified by gene ontology analysis of gene chip data.