基于Backfolding的纯净质谱数据提取算法

Extraction Algorithm of Pure Mass Spectrum Based on Backfolding

混合物样品经过GC-MS系统分析,其中保留时间接近的物质经过色谱后不能被完全分离,导致成分的谱峰重叠,得到的质谱图会包含共流出物的干扰,最终将导致定性分析的准确率和可靠性降低[1]。本文采用基于Backfolding算法提取纯净的质谱谱图。该算法是基于矩阵的算法,首先将原始GC-MS数据表示为矩阵形式,然后将矩阵中相邻扫描质谱图相减,得到的正值和负值的绝对值分别保存在单独的数据集中,相减操作能够去除一些随扫描时间变化较小的噪声干扰。之后,分别对相减得到的两个数据集进行移位操作并重新相加组合成一个矩阵,重新组合的数据的色谱分辨率会有明显的提高[2-4]。最后将新合成的矩阵经过一系列的矩阵运算,可以得到代表纯净成分的质谱数据的矩阵[5-7]。

A novel data analysis algorithm was developed to significantly improve the accuracy and reliability in analysis of a mixture sample by gas chromatography mass spectroscopy（GC-MS）.In the newly-developed algorithm,the pure mass spectra are extracted based on the back-folding method.First,the adjacent scan mass spectra in the raw data matrix are subtracted to remove the noise disturbance little affected by scanning time.Second,after subtraction,the positive data and the absolute values of the negative data are stored in two different data sets,respectively.Last,the two data sets are shifted,summed up,and recombined into a new matrix in such a way that the chromatographic resolution is significantly improved.Moreover,the matrix,representing the pure component mass spectrum,can be derived by appropriate mathematical operations of the recombined matrix.