CFAssay: statistical analysis of the colony formation assay

Area of Science:

• Radiation biology
• Cellular toxicology
• Statistical modeling

Background:

• Colony formation assay is crucial for assessing cell reproductive death post-ionizing radiation.
• Linear-quadratic cell survival curves require robust statistical analysis methods.
• Standardization and code exchange are vital for collaborative research in radiobiology.

Purpose of the Study:

• To introduce CFAssay, an R software package for the statistical analysis of cell survival curves.
• To enable methodological standardization and facilitate code exchange among research laboratories.
• To provide tools for analyzing data from ionizing radiation and chemical treatments.

Main Methods:

• CFAssay implements maximum likelihood methods by default, with optional least squares or weighted least squares.
• Includes statistical tests for comparing cell survival curves and ANOVA for two-way experimental designs.
• Offers a dispersion parameter for goodness-of-fit assessment and various plotting functions.

Main Results:

• Parameter estimates from maximum likelihood and least squares methods show minimal differences in presented examples.
• The dispersion parameter is more sensitive to statistical variation than R-squared from least squares.
• CFAssay facilitates evaluation of experimental data quality.

Conclusions:

• CFAssay enhances the evaluation of experimental data quality through its dispersion parameter and plotting functions.
• As open-source software, CFAssay promotes interlaboratory code sharing and collaboration.
• The package aids in the standardization of linear-quadratic cell survival curve analysis.