基于矩阵分析的STR荧光解谱技术

用多种荧光标记物进行STR检测时，由于荧光光谱的谱带展宽特性，各个荧光光谱之间有重叠部分，如何实现将相互重叠的各个波长上的能量有效利用起来，对提高荧光利用效率至关重要。本文给出了一种基于矩阵分析的数据处理方法，该方法在进行荧光探测前首先要对荧光谱进行光谱校正，即要得到每种所用标记染料受激发射荧光的光谱分布，根据光谱分布建立染料组合荧光信号矩阵，然后对矩阵进行归一，利用归一后的矩阵对所探测的荧光谱进行解谱，从而得到所期望的荧光谱图。理论分析和实验结果表明，该方法可以有效的利用各个波长的荧光光谱能量，并实现对不同荧光重叠谱的有效光谱解谱。

STR Fluorescence Spectral Analysis Based on Matrix Analysis

In a detection system of STR labeled by multiple fluorescence, the overlaps do occur between the different colored fluorophores owing to the band broadening characteristic of the employed fluorescence spectrum. It is crucial to obtain effective energy distribution specific for different wavelengths so as to improve the utilization efficiency of fluorescence intensity. To reach this goal in practice, we propose a data processing method based on matrix analysis. Firstly，a spectral calibration must be done for STR fluorescence spectral analysis, and then a emission spectral distribution of each colored fluorophore excitated by laser is created from the detection signal. And then, the matrix normalization is computed from the above fixed spectrum distribution. And lastly, the inversion of the correct matrix is used to get the resolved and desirable STR spectrogram. In addition, performance evaluation shows that this method based on matrix analysis can make use of the fluorescence spectrum energy of each wavelength and resolve the overlapping spectrum effectively.