Improvements in methods for calculating virus titer estimates from TCID50 and plaque assays
Improvements in methods for calculating virus titer estimates from TCID50 and plaque assays
复制标题
DOI：
10.1016/s0166-0934(01)00316-0
复制
发表时间：
2001-08-01
影响因子：
3.1
通讯作者：
Lowy, RJ
中科院分区：
文献类型：
--
作者：
LaBarre, DD;Lowy, RJ
计算滴度估计值和使用滴度降低试验是病毒学家使用的基本方法。作为生物测定和基于有限稀释法的滴度测定在方法学和分析上都需要良好的误差控制。随着临床、生产和研究环境的采用，对良好的统计分析的需求可能会变得更大，分析标准、质量控制和保证标准也会得到改进。此外，越来越多的病毒滴度测定基于高通量方法，其产生连续的而不是传统的量子数据。本文描述了两种不同的加权线性回归方法，用于测定CPE检测的TCID 50和PFU滴度。TCID 50分析使用了使用连续比色数据的广义最小二乘法。菌斑分析利用加权最小二乘法，通过使用系列稀释产生的量子菌斑数据的原点。与简单的计算方法相比，这两种方法在滴度和误差估计方面都有所改进。当缺乏实验材料或分析成本无法进行广泛的重复滴度测定时，这些方法可能具有最大的价值，但必须对滴度和/或治疗差异进行良好的估计。(C)2001 Elsevier Science B. V.保留所有权利。
Calculation of titer estimates and use of titer reduction assays are fundamental approaches used by virologists. Titer assays being biological assays and based on limiting dilution methods require good error control, both methodologically and analytically. The need for good statistical analysis is likely to become even greater as in clinical, manufacturing, as well as the research settings, improved analytical criteria, quality control, and assurance standards are adopted. Furthermore, increasingly, virus titer assays are based on high throughput methods, which generate continuous rather than traditional quantal data. Described here are two different weighted linear regression methods to determine TCID50 and PFU titers from CPE assays. The TCID50 analysis makes use of a generalized least squares approach using continuous colorimetric data. The plaque analysis makes use of weighted least squares forced through the origin using quantal plaque data generated by serial dilutions. Both methods are improvements in titer and error estimation compared to simpler calculation methods. These methods may have greatest value when lack of experimental material or costs of analysis precludes extensive replicate titer determinations but good estimates of titers and/or treatment differences are essential. (C) 2001 Elsevier Science B.V. All rights reserved.