Nurbaev S D, Balanovskaia E V
Institute of Clinical Genetics, Russian Academy of Medical Sciences, Moscow, Russia.
Genetika. 1998 Jun;34(6):825-38.
An important aspect of gene geography, the estimation of the reliability of interpolation maps, is considered. The introduced quantitative parameter, map reliability, was estimated as the probability of predicted character values in interpolated map regions. The obtained estimates characterize the statistic significance of the mapped values of the character. The estimation algorithm is based on concepts and mathematical methods of the reliability theory. The proposed approach involved the estimation of the reliability at each point of the mapped area and resulted in a new map (a reliability map) expressing the reliability of gene geographic mapping in probability terms. Approaches to estimate the reliability of mapping, as dependent on various parameters of the initial data, were proposed, a general computer-based technology was elaborated, and a standardized reliability scale was proposed. Reliability maps are considered necessary for the correct interpretation of gene geographic maps. The estimation of reliability as dependent on the number and distribution of initially tested populations was illustrated by the example of frequency maps of individual genes (HP1, HLAA1) and synthetic maps (100 alleles of 34 polymorphic loci) of Eastern Europe.
本文考虑了基因地理学的一个重要方面,即插值图可靠性的估计。引入的定量参数——地图可靠性,被估计为插值图区域中预测特征值的概率。所获得的估计值表征了特征映射值的统计显著性。该估计算法基于可靠性理论的概念和数学方法。所提出的方法涉及对映射区域中每个点的可靠性进行估计,并生成一张新地图(可靠性地图),该地图用概率术语表达了基因地理映射的可靠性。提出了根据初始数据的各种参数来估计映射可靠性的方法,阐述了一种通用的基于计算机的技术,并提出了标准化的可靠性量表。可靠性地图被认为是正确解释基因地理图所必需的。以东欧个体基因(HP1、HLAA1)频率图和合成图(34个多态位点的100个等位基因)为例,说明了可靠性估计如何依赖于最初测试群体的数量和分布。