生物信息技术分析脑胶质瘤血清蛋白指纹图诊断模型的临床意义  

Serum protein fingerprinting coupled with artificial neural network distinguishes glioma from healthy population and brain benign tumo

作　　者：刘建[1] 吴敏良[2] 董琦[1] 余捷凯[1] 胡跃[1] 徐文莉[2] 沈宏[1] 郑树[1] 

机构地区：[1]浙江大学附属第二医院肿瘤研究所,杭州310009 [2]浙江大学附属第二医院检验科,杭州310009

出　　处：《中华检验医学杂志》2005年第1期30-33,共4页Chinese Journal of Laboratory Medicine

基　　金：国家 973计划重点基础研究发展规划资助项目(G19980 5 12 0 0 )

摘　　要：目的　探讨筛选的脑胶质瘤血清蛋白指纹图诊断模型的临床应用价值。方法　用疏水性表面芯片 (H4 )和表面增强激光解吸离子化飞行时间质谱技术 (SELDI TOF MS)及生物信息学分析方法与SPSS10 0软件 ,检测分析 2 8例胶质瘤、37例其他脑良性肿瘤和 4 0名健康人的血清蛋白指纹图 ,并建立脑胶质瘤血清蛋白指纹图诊断模型。结果　用建立的区分脑胶质瘤与健康人的血清蛋白指纹图诊断模型进行盲法检测的准确率、敏感性和特异性分别为 95 7% (2 2 / 2 3)、88 9% (8/ 9)和 10 0 % (14 / 14 )。建立的区分胶质瘤与其他脑良性肿瘤的血清蛋白指纹图诊断模型 ,盲法检测脑胶质瘤的准确率、敏感性和特异性分别为 86 4 % (19/ 2 2 )、88 9% (8/ 9)和 84 6 % (11/ 13)。建立的区分Ⅰ Ⅱ级与Ⅲ Ⅳ级脑胶质瘤的血清蛋白指纹图诊断模型的准确率分别为 85 7% (13/ 15 )、84 6 % (11/13)。结论　用SELDI TOF MS技术与生物信息分析法建立的 3个血清蛋白指纹图诊断模型对脑胶质瘤的定性诊断提供了一条新途径。objective To screen and evaluate protein biomarkers for the detection of glioma from healthy population and from brain benign tumor by using surface enhanced laser desorption/ionization mass spectrometry (SELDI MS) coupled with an artificial neural network algorithm (ANN) Methods SELDI TPF MS protein fingerprinting of serum from 105 patients of brain tumors and healthy men, including 28 patients with glioma, 37 patients with brain benign tumor, and 40 age matched healthy men were included in the study Two thirds of samples of total number of every compared pair were used to set up discriminating patterns One third of samples of total number of every compared pair were used to cross validate, simultaneously The discriminate cluster analysis derived SPSS10 0 software was used to compare gradeⅠ Ⅱgliomas with grade Ⅲ Ⅳones Results A blinded test showed a accuracy of 95 7%, a sensitivity of 88 9%, a specificity of 100%, a positive predictive value of 90% and negative predictive value of 100% when comparing the gliomas with healthy men A accuracy of 86 4%, a sensitivity of 88 9%, a specificity of 84 6%, were obtained when comparing the gliomas with benign brain tumors A total accuracy of 85 7%, a accuracy ofⅠ Ⅱgloma was 86 7%, a accuracy of Ⅲ Ⅳgloma was 84 6% when comparing the grade with grade Ⅲ Ⅳones (discriminate cluster analysis) Conclusion SELDI MS combined with bioinformatics tools can greatly facilitate the discovery of new and better biomarkers The high sensitivity and specificity achieved by the use of the selected biomarkers show great potential application for the discrimination of gliomas from healthy person and from brain benign tumors

关 键 词：脑胶质瘤 血清蛋白 指纹图 盲法 TOF 健康人 临床意义 生物信息技术 SELDI 特异性 

分 类 号：R739.4[医药卫生—肿瘤]