Zheng Qi
Department of Epidemiology and Biostatistics, Texas A&M School of Public Health, College Station, TX 77843, United States.
Mutat Res. 2015 Jul;777:20-2. doi: 10.1016/j.mrfmmm.2015.04.002. Epub 2015 Apr 9.
Comparing microbial mutation rates via the fluctuation assay protocol is an important routine task in the laboratory. However, methods for the comparison of mutation rates are still scarce and not well understood. This paper proposes new methods to address this situation. First, it provides a likelihood ratio test. Second, it explores the use of confidence intervals that can be readily computed. Third, it uses simulations to assess the new and existing methods. Both the likelihood ratio test and the use of confidence intervals were found to be superior to the general-purpose Mann-Whitney test, while the t-test was found to be unsuitable for fluctuation assay data. In addition, the method based on confidence intervals is suitable for comparison of experiments in which the terminal cell population sizes differ.
通过波动试验方案比较微生物突变率是实验室中的一项重要日常任务。然而，用于比较突变率的方法仍然稀缺且尚未得到充分理解。本文提出了新的方法来解决这种情况。首先，它提供了一种似然比检验。其次，它探索了可轻松计算的置信区间的使用。第三，它使用模拟来评估新方法和现有方法。发现似然比检验和置信区间的使用均优于通用的曼-惠特尼检验，而t检验被发现不适用于波动试验数据。此外，基于置信区间的方法适用于比较终末细胞群体大小不同的实验。
