CAJ | 학술논문

针对表面增强激光解吸/离子化飞行时间质谱（SELDI-TOF MS）血清多肽谱解析,首先采用离散小波变换（discrete wavelet transform,DWT）对数据进行滤噪、压缩;再采用偏最小二乘（partial leastsquares,PLS）法建立癌症诊断预测模型.考察了DWT对血清多肽谱数据预滤噪、压缩的效率,同时重点研究了小波分解尺度、主成分数对诊断预测准确率的影响.研究结果表明,采用DWT-PLS法解析血清多肽谱简便、高效、预测准确.
침대표면증강격광해흡/리자화비행시간질보（SELDI-TOF MS）혈청다태보해석,수선채용리산소파변환（discrete wavelet transform,DWT）대수거진행려조、압축;재채용편최소이승（partial leastsquares,PLS）법건립암증진단예측모형.고찰료DWT대혈청다태보수거예려조、압축적효솔,동시중점연구료소파분해척도、주성분수대진단예측준학솔적영향.연구결과표명,채용DWT-PLS법해석혈청다태보간편、고효、예측준학.
The SELDI-TOF MS serum peptide profile is dealt through the discrete wavelet transform（DWT） and partial least squares（PLS）.The spectral background and noise are filtered,and the profile was compressed by DWT.Then,the prediction model was constructed by PLS for cancer prognosis.The efficiency of DWT and PLS was studied and the influences of parameters in method were considered.The results indicated that DWT-PLS was convenient,high efficient and accurate.