论文部分内容阅读
目的:应用液体蛋白芯片飞行时间质谱系统分析胃癌患者血清蛋白质表达谱,寻找具有潜在诊断意义的血清标志物。方法:收集血清样本62例,其中正常对照组(N组)16例,胃癌组(T组)28例,验证组18例。经WCX磁珠纯化、MALDI-TOF-MS及ClinproTools生物信息学方法研究其血清蛋白表达谱,并筛选出差异蛋白质峰,运用数据挖掘算法,构建胃癌的血清蛋白诊断模型,并在验证组中验证其准确性。结果:1)通过对比胃癌组和正常组的血清蛋白质谱图,分析得到25个具有显著差异的蛋白质峰,其中差异最显著的前两位质核比分别为5 248.49m/z和5 754.25m/z,其灵敏度分别为84.61%和73.07%,特异性分别为100%和93.75%,能很好地区分胃癌组和正常组。2)通过ANN的数据挖掘的方法,在具有显著差别的25个蛋白质峰中,筛选了组合能力最强的6个蛋白峰(分别为4 268.05m/z、5 636.53m/z、5 248.49m/z、2 933.15m/z、1450.13m/z和1 349.4m/z),建立了胃恶性肿瘤的诊断模型,其识别率为100%,预测能力为90.59%,准确性为100%。将已知信息的验证组18例分别代入已建立的模型,特异性和灵敏性分别为75%和100%。结论:液体蛋白芯片飞行时间质谱系统作为研究蛋白表达谱的工具,能够用于筛选潜在的胃恶性肿瘤的血清标志物,利用其优点并结合统计学的方法,建立血清学胃癌的诊断模型,能为胃恶性肿瘤的筛查提供帮助。
OBJECTIVE: To analyze the serum protein profile of patients with gastric cancer using liquid protein chip time-of-flight mass spectrometry and to search for serum markers with potential diagnostic significance. Methods: Sixty-two serum samples were collected, of which 16 were in normal control group (N group), 28 in gastric cancer group (T group) and 18 in validation group. The protein profile of serum was studied by WCX bead purification, bioinformatics methods of MALDI-TOF-MS and ClinproTools, and the differential protein peaks were screened. The data mining algorithm was used to construct the diagnostic model of serum protein in gastric cancer and validated in the validation group Its accuracy. Results: 1) By comparing the serum protein profile of gastric cancer group and normal group, we obtained 25 protein peaks with significant difference, the most significant difference was 5 248.49 m / z and 5 754.25 m / z, the sensitivity was 84.61% and 73.07% respectively, the specificity was 100% and 93.75% respectively, which could distinguish gastric cancer group from normal group well. 2) By means of data mining by ANN, among the 25 protein peaks with significant difference, six protein peaks with the highest combination ability were screened (4 268.05 m / z, 5 636.53 m / z, 5 248.49 m / z, 2 933.15 m / z, 1450.13 m / z and 1 349.4 m / z), a diagnostic model of gastric malignant tumor was established with a recognition rate of 100%, a prediction ability of 90.59% and an accuracy of 100%. The known information validation group 18 cases were substituted into the established model, specificity and sensitivity were 75% and 100%. Conclusion: The liquid protein chip time of flight mass spectrometry system can be used as a tool to study the protein expression profiling. It can be used to screen serum markers of potential gastric malignancies. Based on its advantages and combined with statistical methods, a diagnostic model of serological gastric cancer can be established. To help diagnose gastric cancer.