中文摘要：

目的研究建立大肠癌（colorectal cancer,CRC））唾液蛋白质指纹图谱新型分子诊断模型。方法采集34例肠癌患者、45例健康志愿者（正常组）的唾液标本,用弱阳离子交换型（WCX）纳米磁珠联合基质辅助激光解析电离飞行时间质谱（matrix assisted laser desorption ionization time-of-flight mass spectrometry,MALDI-TOF MS）技术进行检测,获得各标本的蛋白指纹图谱。用Biomarker Wizard软件分析所获得的蛋白指纹图谱找出差异蛋白,再用Biomarker Patterns 5.0.2建立鉴别诊断模型。结果共检测到312个肠癌差异蛋白质峰,两组比值大于3（肠癌/正常对照〉3,或者正常对照/肠癌〉3）,其中有37个差异蛋白质峰有统计学意义（P〈0.05）;其中有7个差异蛋白质峰肠癌组表达上调,28个差异蛋白质峰肠癌组表达下调,有35个差异蛋白质峰有显著差异（P〈0.01）。筛选建立了由m/z为2 501.26、4 779.95、3 140.39的3个差异蛋白峰组成的肠癌与正常组的诊断模型,该模型的灵敏度和特异度分别为88%（30/34）和98%（44/45）;通过交叉验证法进一步验证诊断模型的可靠性,结果该模型的灵敏度和特异度分别为85%（29/34）和88%（37/45）。结论用WCX结合MALDI-TOF-MS技术建立的唾液蛋白诊断模型为肠癌的诊断提供了新途径。

英文摘要：

Objective To establish a novel molecular diagnostic model of saliva protein fingerprint in colorectal cancer（CRC） patients. Methods Saliva samples from 34 patients with CRC, and 45 healthy people were analyzed by weak cationicexchange magnetic beads（MB-WCX） and matrix-assisted laser desorption ionization time-of-flight mass spectrometry（MALDITOF-MS） methods. Subsequently, we compared the saliva peptide signatures of the two groups and obtained differently expressed peptides by using of Biomarker Wizard, then establish a diagnostic model to diagnose gastric carcinoma by using of Biomarker Patterns 5.0.2. Results 312 differentially expressed protein peaks were detected, the ratio of two groups 3.0（CRC/control〉3.0, or control/ CRC〉3.0）, including 37 protein peaks were statistically significant（P〈0.05）; 7 peaks were upregulated, 28 peaks were down-regulated, there were 35 different protein peaks have significant difference（P〈0.01）.Further more, we screened and built a saliva proteomic models with 3 protein molecules m/z 2501.26, 4779.95, 3140.39 to distinguish CRC groups and normal groups. The sensitivity of this model was 88%（30/34）, and the specificity was 98 %（44/45）. The reliability of this model was further verified with a sensitivity of 85%（29/34） and a specificity of 88%（37/45） by cross validation method. Conclusion Saliva proteomic profiling by using MALDI-TOF-MS combined with WCX technique is a novel potential tool for the clinical diagnosis of CRC.