Kravatskiĭ Iu V, Poletaev A I
Mol Biol (Mosk). 1994 Jul-Aug;28(4):887-99.
The proposed procedure of computer analysis of the flow karyotype data, obtained in human chromosomes studies, is able to provide information about the basic parameters of the karyotypes: the positions of the peaks (corresponding to the relative size of chromosomes), peaks areas (relative number of chromosomes in the sample), coefficients of variation (CV) of the peaks--possible differences between homologous chromosomes. The analysis is based on the assumption that all chromosomal components of the experimental distributions are normal (Gaussians). The algorithm of the analysis uses a combination of two approaches: truncation method and least squares method. As the flow data are "contaminated" by background components, special tools for filtering off the contaminating signals were designed including the original integral Fourier filtering procedure. This analysis is realized in a program package utilizing IBM-compatible PCs. The user is able to get the desired parameters for most chromosomes of the karyotype under study from univariate flow data: differences between particular homologous chromosomes, presence of chromosome aberrations, extra chromosomes, etc., since structural aberrations and chromosome number variation lead to specific changes of the parameters of chromosome-related components.
所提出的用于对人类染色体研究中获得的流式核型数据进行计算机分析的程序，能够提供有关核型基本参数的信息：峰的位置（对应于染色体的相对大小）、峰面积（样本中染色体的相对数量）、峰的变异系数（CV）——同源染色体之间可能存在的差异。该分析基于实验分布的所有染色体成分均呈正态分布（高斯分布）这一假设。分析算法采用了两种方法的组合：截断法和最小二乘法。由于流式数据被背景成分“污染”，因此设计了特殊的工具来滤除污染信号，包括原始的积分傅里叶滤波程序。这种分析在一个利用IBM兼容个人计算机的程序包中实现。用户能够从单变量流式数据中获取所研究核型的大多数染色体的所需参数：特定同源染色体之间的差异、染色体畸变的存在、额外染色体等，因为结构畸变和染色体数量变异会导致与染色体相关成分的参数发生特定变化。
