Li Zhen, Shang Li-Ping, Deng Hu, Zhi Tong-Xing
College of Information Engineering, Southwest University of Science and Technology, Mianyang 621010, China.
Guang Pu Xue Yu Guang Pu Fen Xi. 2009 Jul;29(7):1925-8.
In the present paper, three-dimensional fluorescence spectra of tyrosine and tryptophan were detected through Pekin-Elmer LS55 fluorescence spectrometer made in U. S. A, with the excitation wavelength in the range of 230-320 nm, interval of 5 nm, emission wavelength in the range of 230-500 nm, and interval of 2 nm. Three-dimensional fluorogram was obtained using excitation wavelength, emission wavelength and fluorescence intensity detected by Pekin-Elmer LS55 as threedimensional system of coordinate. It can be seen that the main peak of the three-dimensional fluorescence spectra of tyrosine and tryptophan was aliasing evidently, so it was impossible to distinguish these two overlapping components using excitation wavelength, emission wavelength and fluorescence intensity easily. Therefore, to distinguish spectra of tyrosine and tryptophan obviously, characteristic parameter was extracted based on the principle of mathematical statistics and the most relevant information of the fluorescence spectra of tyrosine and tryptophan was obtained. The results showed that the difference value percentage of the "mean", "deviation", "origin pitch" and "mixed center pitch" of the three-dimensional spectra' s characteristic parameter was 330.37%, 102.86%, 329.16% and 329.63%, respectively; meanwhile, the difference value percentage of "distribution" and "correlation" was 10.61% and 2.40% respectively. Thus, it was obvious that the "mean", "deviation", "origin pitch" and "mixed center pitch" of the three-dimensional spectra's characteristic parameter can be used to distinguish overlap spectra of tyrosine and tryptophan as sensitive characteristic parameters. The principle and results of this method are applicable and valid. This "math pre-extracted" three-dimensional spectrometry found sensitive parameter among the components by the parameter extracting and can replace traditional excitation-emission-intensity three-dimensional fluorescence analysis, then can be generalized to the identification of multi-components.
在本文中，使用美国珀金埃尔默公司的LS55荧光光谱仪检测酪氨酸和色氨酸的三维荧光光谱，激发波长范围为230 - 320nm，间隔为5nm，发射波长范围为230 - 500nm，间隔为2nm。以珀金埃尔默LS55检测到的激发波长、发射波长和荧光强度作为三维坐标系，获得三维荧光图谱。可以看出，酪氨酸和色氨酸的三维荧光光谱主峰明显重叠，因此难以通过激发波长、发射波长和荧光强度轻易区分这两个重叠成分。所以，为了明显区分酪氨酸和色氨酸的光谱，基于数理统计原理提取特征参数，获得了酪氨酸和色氨酸荧光光谱最相关的信息。结果表明，三维光谱特征参数的“均值”“偏差”“原点间距”和“混合中心间距”的差值百分比分别为330.37%、102.86%、329.16%和329.63%；同时，“分布”和“相关性”的差值百分比分别为10.61%和2.40%。因此，很明显三维光谱特征参数的“均值”“偏差”“原点间距”和“混合中心间距”可作为敏感特征参数用于区分酪氨酸和色氨酸的重叠光谱。该方法的原理和结果是适用且有效的。这种“数学预提取”三维光谱法通过参数提取在各成分中找到敏感参数，可取代传统的激发 - 发射 - 强度三维荧光分析，进而可推广到多成分的鉴别。
