CAJ | 학술논문

目的比较骨肉瘤患者和正常对照者血清蛋白表达谱的差异,筛选骨肉瘤相关血清蛋白标志物,并建立基于决策树的预测模型,为筛选和建立骨肉瘤临床诊断的血清学指标提供依据.方法 27例骨肉瘤患者血清(男17例,女10例)及47名相匹配者正常对照血清标本随机分为两组:60份(23例骨肉瘤,37名正常对照)为建模组,14份(4例骨肉瘤,10名正常对照)为盲法筛选组.利用表面增强激光解吸离子化-飞行时间-质谱(surface enhanced laser desorption/ionization time of fight mass spectrometry,SELDI-TOF-MS)技术进行蛋白质谱分析.采用蛋白质飞行质谱仪对结合在CM10芯片上的血清蛋白进行读取分析.通过Biomarker Wizard软件比较两组人群血清蛋白质谱的差异,经生物信息学分析得到决策树模型并进行盲法验证.结果在质荷比(M/Z)1488.15～19842u范围内,共检测到96个有效蛋白峰,其中9个峰差异有统计学意义.利用三倍交叉证实方法对建模组的蛋白质谱数据进行1000次随机抽样,得到1000个决策树.根据交叉证实的正确率选出最佳的20个决策树模型作为最终预测模型.用其对14个盲法筛选样本进行归类预测的正确率为85.71%.结论应用SELDI-TOF-MS技术可筛选出骨肉瘤相关血清蛋白标志;建立的决策树模型可以对骨肉瘤作出较为准确的预测判断.
목적 비교골육류환자화정상대조자혈청단백표체보적차이,사선골육류상관혈청단백표지물,병건립기우결책수적예측모형,위사선화건립골육류림상진단적혈청학지표제공의거.방법 27례골육류환자혈청(남17례,녀10례)급47명상필배자정상대조혈청표본수궤분위량조:60빈(23례골육류,37명정상대조)위건모조,14빈(4례골육류,10명정상대조)위맹법사선조.이용표면증강격광해흡리자화-비행시간-질보(surface enhanced laser desorption/ionization time of fight mass spectrometry,SELDI-TOF-MS)기술진행단백질보분석.채용단백질비행질보의대결합재CM10심편상적혈청단백진행독취분석.통과Biomarker Wizard연건비교량조인군혈청단백질보적차이,경생물신식학분석득도결책수모형병진행맹법험증.결과 재질하비(M/Z)1488.15～19842u범위내,공검측도96개유효단백봉,기중9개봉차이유통계학의의.이용삼배교차증실방법대건모조적단백질보수거진행1000차수궤추양,득도1000개결책수.근거교차증실적정학솔선출최가적20개결책수모형작위최종예측모형.용기대14개맹법사선양본진행귀류예측적정학솔위85.71%.결론 응용SELDI-TOF-MS기술가사선출골육류상관혈청단백표지;건립적결책수모형가이대골육류작출교위준학적예측판단.
Objective To provide some theoretic evidence for screening and establishing serum in dieators of early diagnosis of osteosarcoma(OS),the serum proteomics profiling difference of subjects with OS and age-matched healthy controls were analyzed to screen senlm proteomic biomarker related to osteosarcoma.Methods Serum samples were collected from 27 patients of OS (17 males,10 females) and 47 age and sex-matched healthy controls.The samples were divided into 2 sets randomly:training set (23 OS patients,37 healthy controls) and blind testing set(4 OS patients,10 healthy controls).Special serum protein or peptide pattern was determined by SELDI-TOF-MS measurement after treating the sample onto CM10 protein chip.The obtained data were analyzed by Biomarker Wizard software to screen serum proteome biomarkers with relation to OS.while decision tree for diagnosis of OS and blind validation were determined by bioinformatics analysis.Results 96 effective protein peaks were detected at the molecular range of 1488.15-19842u,among which 9 were significantly different between OS and controls.All tlle peptide pattern data were sampled randomly 1000 time.and 1000 decision tree model were obtained.Decision tree and 3-cross validation approach were used combine,20 decision tree which can difierentiate effectively OS patients from controls were constructed.With these classification tree.12 samples were correctly forecasted in 14 blind testing samples.the corresponding correct rate was 85.71%.Conclusion SELDI-TOF-MS protein chip combined with artificial intelligence classification algorithm helps find serum proteome biomarkers related to OS and the predictive medels can discriminate OS from healthy controls effectively,which may have gome potential value for early diagnosis of OS.