慢性肾功能衰竭中医湿证舌苔中相关蛋白研究

目的：探讨慢性肾衰舌苔上清液中相关蛋白变化与慢性肾衰中医湿证的相关性。方法：临床收集慢性肾衰患者舌苔样本102例,其中湿证组76例,非湿证组26例;另有正常对照组38例,对所有样本采用1500r/min的离心转速,离心2min后取上清液,选用NP20蛋白芯片点样,经SELDI-TOF-MS技术测定得到蛋白质谱峰,通过Biomarker Wizard软件比较湿证组、非湿证组及正常对照组之间蛋白质谱峰的差异,利用层次聚类算法进行层次聚类分析和主成分分析（PCA）,经生物信息学分析得到慢性肾衰中医湿证决策树模型。结果：（1）慢性肾衰湿证76例、非湿证26例及正常对照组38例舌苔样本经SELDI-TOF-MS技术测定,质荷比1000～20000范围内共检测到136个蛋白峰,经生物信息学统计分析,有11个差异质谱峰有统计学意义（P〈0.05）,其中,M/Z2434.53、M/Z3672.89、M/Z7118.00和M/Z8411.134个差异质谱峰在湿证组和非湿证组呈高表达,M/Z3333.52、M/Z4619.83、M/Z4711.35和M/Z10664.24个差异质谱峰在湿证组和非湿证组呈低表达,M/Z5142.52、M/Z5231.98和M/Z5272.633个差异质谱峰在湿证组呈高表达,在非湿证组呈低表达;（2）利用层次聚类算法进行层次聚类分析和主成分分析（PCA）,慢性肾衰湿证组、非湿证组与正常对照组两两之间均有明显的区分度,但都存在部分重叠;（3）经生物信息学分析建立慢性肾衰中医湿证决策树模型,最终得到M/Z5231.98、M/Z2325.41、M/Z4558.86和M/Z3491.894个差异质谱峰组成的生物标记物可以将湿证组和非湿证组样品较好的分类（最终决策树模型对所有样品分组判别的正确率为83.33%,灵敏度为81.58%,特异性为88.46%）。结论：该研究从生物学角度来探讨中医湿证,为中医湿证辨证提供客观依据。

The Study on Associated-proteinum of Tongue Coating of Chronic Renal Failure in Damp Syndrome of TCM

Objective：To study the relationship between change of associated-proteinum in tongue coating of CRF and damp syndrome in TCM of CRF. Methods：102 tongue coating samples of CRF patients were collected, among the total samples, the damp syndrome group has 76 samples, the non-damp syndrome group has 26 samples, and normal control group has 38 samples. All samples were centrifuged at 1500r/min, after 2min, dislodged clear supernatant liquid and spoted one by one to proteinum chips of NP20, detected by technique of SELDI-TOF-MS for proteinum peaks, contrasted by software named Biomarker Wizard for the differences of proteinum peaks among damp syndrome group, non-damp syndrome group and normal control group, utilized fuzzy grouping algorithm to run fuzzy grouping analysis and principal component analysis, analysised by bioinformatics for decision tree model of damp syndrome in TCM of CRF.Results：（1）Tongue coating samples of CRF of 76 damp syndrome group, 26 non-damp syndrome group and 38 normal control group are detcted by technique of SELDI-TOF-MS, 136 proteinum peaks have been detected at 1000～20000 e/m, 11 distincted mass spectrum peaks have been analysised by bioinformatics, which have statistical significance （P0.05）, the four M/Z2434.53, M/Z3672.89, M/Z7118.00 and M/Z8411.13 show high expression in damp syndrome group and non-damp syndrome group, the four M/Z3333.52, M/Z4619.83, M/Z4711.35 and M/Z10664.2 show low expression in damp syndrome group and non-damp syndrome group, the three M/Z5142.52, M/Z5231.98 and M/Z5272.63 show high expression in damp syndrome group and low expression in non-damp syndrome group.（2）Using fuzzy grouping algorithm to run fuzzy grouping analysis and principal component analysis, among damp syndrome group, non-damp syndrome group and normal control group, there have conspicuous discriminations. （3）Analyzed and established decision tree model of damp syndrome in TCM of CRF by bioinformatics, finally, we obtain the biological markers constituted by 4 distincted mass spectrum peaks of M/Z5231.98,M/Z2325.41, M/Z4558.86 and M/Z3491.89 can distinguish well between damp syndrome group samples and non-damp syndrome group samples （The decision tree model group-discriminates to all samples, whose exactitude rate is 83.33%, sensitivity is 81.58% and specificity is 88.46%）. Conclusion：The study explores the damp syndrome in TCM on the point of biology, which provides objective evidence for differentiation of symptoms and signs of damp syndrome in TCM.