Evaluating statistical analyses and reproducibility of microbial mutagenicity assays

Area of Science:

• Toxicology
• Genetics
• Biostatistics

Background:

• The National Cancer Institute/National Toxicology Program (NCI/NTP) conducted a multi-laboratory study to assess the reproducibility of the Salmonella/microsome assay.
• This assay is a standard method for evaluating chemical mutagenicity.

Purpose of the Study:

• To perform a statistical analysis of data generated from the Salmonella/microsome assay.
• To evaluate the reproducibility and decision-making adequacy of various statistical approaches for determining chemical mutagenicity.

Main Methods:

• Data cleaning involved identifying and removing spurious results.
• Seven distinct statistical methods were applied: 2-fold rule, modified 2-fold rule, ANOVA, linear trend test, combined ANOVA and trend test, 97.5th percentile threshold, and confidence interval threshold.
• Statistical conclusions were compared against microbiologists' interpretations.

Main Results:

• The study compared the conclusions drawn by each of the seven statistical rules against expert judgment.
• Strengths and weaknesses of each statistical approach were discussed in the context of mutagenicity assessment.
• The overall reproducibility of the Salmonella/microsome assay was examined.

Conclusions:

• The statistical analysis provided insights into the reliability of different methods for interpreting mutagenicity data.
• The findings contribute to understanding the robustness of the Salmonella/microsome assay and inform best practices for data analysis.