基于蛋白质谱技术建立胃癌诊断预测模型及其初步验证  

Establishment of diagnostic and prediction model of gastric cancer and its preliminary verification based on protein mass spectrometry

作　　者：王娜娜[1,2] 张婷 宋御繁[1,2] 胡成进 陈英剑 WANG Nana;ZHANG Ting;SONG Yufan;HU Chengjin;CHEN Yingjian(Taishan Medical University,Taian,Shandong 271016,China;Weifang Medical University,Wei fang,Shandong 261053,China;Department of Laboratory Diagnosis,People＇s Liberation Army No.960 Hospital,Jinan,Shandong 250031,China)

机构地区：[1]泰山医学院,山东泰安271016 [2]解放军第九六○医院实验诊断科,山东济南250031 [3]潍坊医学院,山东潍坊261053

出　　处：《国际检验医学杂志》2018年第22期2741-2745,共5页International Journal of Laboratory Medicine

摘　　要：目的基于高通量蛋白质谱技术利用ClinProTools系统筛选胃癌患者与健康对照者的血清差异蛋白/多肽,建立胃癌诊断预测模型,发现胃癌的潜在血清标志物。方法收集胃癌患者血清标本60例、对照组血清标本60例,按照3∶1随机分为训练组(胃癌患者45例,健康对照45例)和验证组(患者15例,健康对照15例)。采用弱阳离子交换磁珠(WCX-MB)吸附低丰度蛋白/多肽,通过基质辅助激光解吸电离飞行时间质谱(MALDI-TOF MS)分析建立相应的蛋白表达指纹图谱,筛选差异表达蛋白/多肽,建立胃癌的诊断预测模型并进行验证。结果筛选出10个具有显著差异的多肽峰,差异最显著的2个多肽峰质荷比(m/z)2 952.57和m/z 5 904.36的受试者工作曲线(ROC)曲线下面积(AUC)分别为0.93和0.81。运用遗传算法(GA)建立胃癌诊断预测模型,验证组验证此模型,结果显示此模型准确性为90.00%,灵敏度为93.30%,特异度为86.70%。结论 MALDI-TOF MS联合WCX-MB能够检测到胃癌患者与健康对照者的血清差异蛋白/多肽,以此建立的胃癌蛋白质谱诊断预测模型存在着潜在的应用价值,m/z 2 952.57、5 904.36有望成为诊断胃癌的潜在血清标志物。Objective To screen the differential proteins/peptides of gastric cancer patients and healthy controls by ClinProTools system based on high throughput protein mass spectrometry, and to establish the gastric cancer diagnosed prediction model for founding the potential serum markers of gastric cancer. Methods 60 patients with gastric cancer and 60 healthy controls were randomly divided into the training group （45 patients, 45 healthy controls） and the validation group （15 patients,15 healthy controls）. Low abundance pro teins/peptides were adsorbed by weak cation exchange magnetic beads （WCXMB）,and the corresponding protein expression fingerprints were established by matrix assisted laser desorption/ ionization time of flight mass spectrometry （MALDI TOF MS） to screen differentially expressed proteins/peptides, established and validated the diagnostic prediction model for gastric cancer. Results A total of 10 significant differential pep tide peaks were screened,Areas under the curve （AUC） in receiver operating characteristics （ROC） for the two most significant difference peaks of mass to charge ratio （m/z） 2 952.57 and m/z 5 904.36 were 0.93 and 0.81, respectively. The diagnostic prediction model of gastric cancer was established by using the genetic algo rithm （GA）. The validation group verified the model. The results showed that the accuracy of the model was90.00%,the sensitivity and the specificity were 93.30% and 86. 70% , respectively. Conclusion MALDI TOF MS combined with WCX MB can detect the serum differential proteins/peptides of gastric cancer patients and healthy controls,and the established gastric cancer protein prognosis prediction model has potential clinical ap plication value,m/z 2 952.57,5 904.36 might be the potential diagnostic marker for gastric cancer.

关 键 词：基质辅助激光解吸电离飞行时间质谱 蛋白质组学 胃癌 诊断预测模型 弱阳离子交换磁珠 

分 类 号：R735.2[医药卫生—肿瘤]