Statistical design and analysis of mutation studies in transgenic mice

Area of Science:

• Genetics and Genomics
• Toxicology and Pharmacology
• Biostatistics

Background:

• Transgenic mouse mutation assays are vital for genetic toxicology studies.
• Significant animal-to-animal variability necessitates robust statistical methods.
• Existing statistical tests may not adequately account for this variability.

Purpose of the Study:

• To identify sources of variability in transgenic mouse mutation assay data.
• To develop appropriate statistical methods and experimental designs for routine studies.
• To contrast the performance of the Cochran-Armitage (CA) test and generalized Cochran-Armitage (GCA) test.

Main Methods:

• Computer simulations were employed to quantify differences between the CA and GCA tests.
• Statistical power and type I error rates were evaluated for each test.
• Simulations were extended to identify cost-effective experimental designs.

Main Results:

• The GCA test performed poorly compared to the CA test when the transgene is the experimental unit and animal numbers are small.
• The CA test performed poorly in small group-size studies when the animal is the appropriate experimental unit.
• Optimal experimental designs can be derived without substantial cost increases.

Conclusions:

• The choice of statistical test (CA vs. GCA) depends on the experimental unit and study size.
• Accounting for animal-to-animal variability is critical for accurate hypothesis testing.
• The study provides a framework for optimizing experimental designs in mutation studies.