Dölle J, Hausmann M, Cremer C
Institute of Applied Physics, University of Heidelberg, FRG.
Anal Cell Pathol. 1991 Mar;3(2):119-32.
Flow cytometry has become a fast, quantitative method for the classification of metaphase chromosomes in suspension (flow karyotyping) stained with fluorescent dyes. Such a flow karyotype (frequency distribution of the fluorescence signals) consists of several peaks. The peak pattern characterizes the analyzed chromosome complement. In many cases flow karyotypes contain a continuum of an unspecific background deriving from chromosome fragments or chromosome aggregates. For the quantitative evaluation of a flow karyotype this background has to be subtracted by a suitable background function. In this approach the application of chi 2-functions is described. The feasibility of this method to flow karyotypes has been concluded from a computer simulation of chromosome breaking under different conditions. In spite of the rather rough assumptions of the model compared to the complex reasons that influence chromosome breaking, the chi 2-function fits the background better than the exponential function in current use. The approximation of a Gaussian distribution function by the chi 2-function also makes it possible to use the same subtraction procedure for chromosome aggregates. The procedure was tested for isolated chromosomes of Chinese hamster cell lines under different states of breaking. For further evaluation of one parameter flow karyotypes a setup of computer routines has been developed for PC/AT and compatible computer systems. Different peak values of these flow karyotypes can be determined (e.g. peak mean, standard deviation, absolute and relative peak area etc.). The applied method is to fit Gaussian curves to each peak of an experimentally measured histogram by using an interactive program. Fluctuations depending on 'noise' may be suppressed by a 'k-nearest-neighbours' smoothing procedure.
流式细胞术已成为一种快速、定量的方法，用于对悬浮状态下用荧光染料染色的中期染色体进行分类（流式核型分析）。这样的流式核型（荧光信号的频率分布）由几个峰组成。峰型表征了所分析的染色体组。在许多情况下，流式核型包含源自染色体片段或染色体聚集体的非特异性背景的连续分布。为了对流式核型进行定量评估，必须通过合适的背景函数减去此背景。在这种方法中，描述了卡方函数的应用。该方法应用于流式核型的可行性已从不同条件下染色体断裂的计算机模拟中得出。尽管与影响染色体断裂的复杂原因相比，该模型的假设相当粗略，但卡方函数比目前使用的指数函数更能拟合背景。卡方函数对高斯分布函数的近似也使得对染色体聚集体可以使用相同的减法程序。该程序在不同断裂状态下对中国仓鼠细胞系的分离染色体进行了测试。为了进一步评估单参数流式核型，已为PC/AT及兼容计算机系统开发了一套计算机程序。可以确定这些流式核型的不同峰值（例如峰均值、标准差、绝对和相对峰面积等）。所应用的方法是通过使用交互式程序将高斯曲线拟合到实验测量直方图的每个峰上。依赖于“噪声”的波动可以通过“k近邻”平滑程序来抑制。
