血清蛋白芯片表达谱在DVT疾病法医鉴定中的应用研究  

Application of serum protein chip expression profile in forensic identification of deep venous thrombosis

作　　者：田英杰 靳茜茜[1] 王一飞 李健 王贵明[2] 牛蕾蕾[1] 曹洁[1] 杜秋香[1] 王英元[1] 孙俊红[1] Tian Yingjie;Jin Qianqian;Wang Yifei;Li Jian;Wang Guiming;Niu Leilei;Cao Jie;Du Qiuxiang;Wang Yingyuan;Sun Junhong(School of Forensic Medicine,Shanxi Medical University,Jinzhong 030600,China;Surgical department,The first hospital of Shanxi Medical University,Taiyuan 030000,China)

机构地区：[1]山西医科大学法医学院,山西晋中030600 [2]山西医科大学第一医院普通外科,山西太原030000

出　　处：《中国法医学杂志》2021年第3期257-262,共6页Chinese Journal of Forensic Medicine

基　　金：国家自然基金面上项目(81470088);山西省青年科技研究基金面上青年基金项目(201901D211334);山西省应用基础研究面上自然基金项目(201701D121175)。

摘　　要：目的利用蛋白芯片技术获取人体和大鼠深静脉血栓(deep venous thrombosis,DVT)组与对照组血清蛋白表达谱的蛋白峰差异,探讨蛋白表达谱与深静脉血栓形成的相关性,为DVT临床和法医学诊断提供依据。方法采用蛋白芯片技术检测DVT患者和正常人(n=10)的血清样本,获得血清蛋白芯片表达谱。建立标准化SD大鼠下腔静脉血栓模型,大鼠被随机分为DVT组和对照组(n=10),蛋白芯片技术检测获得各组血清蛋白表达谱;采用SIMCA-P 14.1软件的正交偏最小二乘判别(orthogonal partial least square-discriminant analysis,OPLS-DA)模型和Python软件的随机森林(Random forest,RF)算法模型对人体和大鼠数据进行判别分析。结果DVT患者和正常人血清蛋白表达峰存在差异,OPLS-DA模型对DVT组和对照组准确判别,R^(2)X=0.631,R^(2)Y=0.928,Q^(2)=0.796;动物模型DVT组与对照组的血清蛋白表达峰亦存在差异,OPLS-DA模型可以将DVT组和对照组准确判别,R^(2)X=0.735,R^(2)Y=0.953,Q^(2)=0.900;人体和动物模型血清中共有8个分子量相似的血清蛋白峰,RF算法模型准确率为100%,人体预测准确率为65%。结论蛋白芯片技术可以快速获取DVT人体和大鼠血清蛋白表达谱,建立OPLS-DA模型和RF模型能对DVT组和对照组进行分类预测分析,为DVT疾病诊断和法医学鉴定提供新的辅助方法。Objective To obtain the serum protein expression profiles between deep venous thrombosis(DVT)group and control group in human and rats by the protein chip technology,to explore the correlation between protein expression profile and DVT,so as to provide basis for clinical and forensic diagnosis of DVT.Methods The protein chip technology was used to detect serum samples of DVT patients and normal people(n=10)to obtain serum protein chip expression profiles.A standardized model of inferior vena cava thrombosis in SD rats was built.Rats were randomly divided into DVT group and control group(n=10).The serum protein expression profile was obtained by protein chip technology;The human and rat data were analyzed by orthogonal partial least squares-discriminant analysis(OPLS-DA)model of SIMCA-P 14.1 software and random forest(RF)algorithm model of Python software.Results The results showed that there were differences in peak of serum protein expression between DVT patients and normal people.OPLS-DA model accurately identified DVT group and control group,R^(2) X=0.631,R^(2) Y=0.928,Q^(2)=0.796.There were differences in peak of serum protein expression between DVT group and normal group in rats.OPLS-DA model accurately identified DVT group and control group,R^(2) X=0.735,R^(2) Y=0.953,Q^(2)=0.900.Research showed that there were eight serum protein peaks with similar molecular weight between human and rat in DVT group.The accuracy of RF algorithm model is 100%,and the accuracy of human prediction is 65%.Conclusion Protein chip technology can quickly obtain the protein expression profile of DVT human and rat serum.OPLS-DA model and RF model can classify and predict DVT group from control group,which provides a new auxiliary method for DVT diagnosis and forensic identification.

关 键 词：法医学 深静脉血栓 Agilent 2100 正交偏最小二乘判别分析 

分 类 号：D918.93[政治法律—法学]