Gubler Hanspeter
NIBR IT and Automation Services, Novartis Institutes for BioMedical Research (NIBR), Basel, Switzerland.
Methods Mol Biol. 2009;552:79-95. doi: 10.1007/978-1-60327-317-6_6.
An overview of the characteristics of classical and outlier-resistant data summaries is provided. The latter are important because outlier data can skew results and decisions based on them. The simple data summaries are the basis for all composite assay and screening data quality measures, for example, the signal-to-noise ratio, signal-to-background ratio, assay and screening window coefficients Z ' and Z, or strictly standardized mean difference (SSMD). In addition to the measures of assay reliability which are based on assessing the size of the "signal windows," some measures for the characterization of the degree of agreement of repeated measurements are also outlined.
本文概述了经典数据汇总和抗异常值数据汇总的特点。后者很重要,因为异常值数据会扭曲基于它们得出的结果和决策。简单的数据汇总为所有复合分析和筛选数据质量指标奠定了基础,例如信噪比、信号与背景比、分析和筛选窗口系数Z'和Z,或严格标准化均差(SSMD)。除了基于评估“信号窗口”大小的分析可靠性指标外,还概述了一些用于表征重复测量一致性程度的指标。