中文摘要：

目的应用蛋白组学技术对前列腺癌血清差异蛋白进行筛选，并建立前列腺癌诊断模型。方法2010年3月至5月间于北京协和医院泌尿外科病房采集12例前列腺癌患者及11例非前列腺癌对照者血清样本，应用弱阳离子磁珠联合基质辅助激光解吸电离飞行时间质谱（matrix—assisted laser desorption／ionization time of flight mass spectrometry，MALDI-TOF MS）技术对前列腺癌患者和对照者血清进行差异蛋白研究，筛选出多个前列腺癌血清差异蛋白，并应用ClinPro Tools2．2软件通过遗传算法建立诊断模型。结果共筛选出前列腺癌与对照组之间差异蛋白峰126个，具有较明显差异的蛋白峰24个（P=0．178）。通过遗传算法优化选择，筛选出符合条件的15个差异蛋白峰建立诊断模型，交叉验证准确性为81．82％，识别能力为100％。结论应用磁珠联合MALDI—TOFMS技术及遗传算法成功建立前列腺癌血清差异蛋白的诊断模型，该模型识别能力高，有助于减少前列腺癌的漏诊率。

英文摘要：

Objective To detect differentially expressed proteins in serum of prostate cancer and to establish a diagnostic model for prostate cancer. Methods Serum samples from 12 patients with prostate cancer and 11 controls hospitalized at our department from March to May 2010 were collected. The weak cation exchange （WCX） beads combined matrix-assisted laser desorption/ionization time of flight mass spectrometry （MALDI-TOF MS） technique was used to detect the differentially expressed proteins in the serum samples of prostate cancer patients and controls. Differentially expressed proteins for prostate cancer were then screened out. Genetic algorithm was utilized to establish a diagnostic model for prostate cancer with ClinProTools 2.2 software. Results Totally 126 different proteins were screened out, of which 24 were significantly different between the prostate cancer patients and controls （P = 0. 178 ）. Using the genetic algorithm, 15 differentially expressed proteins were screened out to establish a diagnostic model. The cross-validation of the model was 81.82% and the recognition rate was 100%. Conclusions A diagnostic model of prostate cancer using the WCX magnetic beads combined MALDI-TOF MS technique and genetic algorithm was successfully established. This model has high cross-validation and recognition rate and is helpful to reduce the misdiagnosis rate of prostate cancer.