摘要
目的:探讨用蛋白质芯片技术筛选多囊卵巢综合征(polycysticovarysyndrome,PCOS)患者血清蛋白质表达谱,寻找血清中的标志性蛋白。方法:采用表面增强激光解离飞行时间质谱技术(surfaceenhancedlaserdesorption/ionizationtimeofflightmassspectrometry,SELDITOFMS),运用WCX2(weakcationexchange)蛋白质芯片检测31例PCOS患者和30例正常对照血清蛋白质谱,获得的蛋白质谱采用BiomarkerWizard软件分析,初步筛选蛋白质峰,结合生物信息学的支持向量机(supportvectormachines,SVM)方法建立并测试PCOS患者的蛋白质指纹图谱模型。结果:在芯片上捕获到225种蛋白质,用质谱仪筛选出PCOS患者与正常对照组相比的23种差异蛋白,从中再次筛选出4种蛋白质组成PCOS的蛋白质谱最优化模型,PCOS患者中质荷比(m/z)分别为6628,6430,6834的3种蛋白质表达上调,m/z为3954的蛋白质表达下调。模型经三倍交叉验证后用盲法测定,其敏感性和特异性分别为83.3%和86.7%,阳性预测值为87.2%。结论:蛋白质芯片技术可以快速、有效地筛选出PCOS患者血清差异蛋白,结合SVM可建立一个由4种蛋白质组成的蛋白质指纹图谱模型,可对PCOS做很好的诊断预测,对这4种蛋白质尤其是m/z为6628的蛋白质进行研究,有助于PCOS病因学进展及诊断标记物的发现。
Objective ： To screen out serum protein profiling of polycystic ovary syndrome （PCOS） by surface - enhancedlaser desorption/ionization time of flight mass spectrometry （ SELDI-TOF MS） for discovering the discriminatory proteins. Methods： Thirty one women with PCOS and thirty healthy women were detected by Weak Cation Exchange（ WCX2 ）chip. ProteinChip reader and Biomarker Wizard software from Ciphergen Inc were combined with a bioinformatics method （support vector machines, SVM ） to analyze protein fingerprinting. Results： There were 4 proteins which were obviously different between the PCOS group and the control. Three of them were up-regulated, and one down-regulated. To set up a analysis model by SVM using the 4 proteins could successfully distinguish between PCOS and the normal control. The corresponding sensitivity, specificity and positive predict value were 86.7%, 83.3% , 87.2%, resectively. Conclusion： Using ProteinChip technology can screen out serum discriminatory proteins quickly and efficiently. Combined with SVM, an optimal fingerprinting model has been set up, which can easily predict PCOS. In disease state of PCOS, there are significant variations which consist of four proteins. To investigate those four discriminatory proteins, especially the protein m/z 6 628 may be of benefit to pathogenic study and the development of biomarkers for PCOS.
出处
《北京大学学报（医学版）》
CAS
CSCD
北大核心
2005年第4期362-365,共4页
Journal of Peking University:Health Sciences
关键词
多囊卵巢综合征
蛋白质阵列分析
光谱分析
质量
Polycystic ovary syndrome
Protein array analysis
Spectrum analysis, mass
