TestDimorph: An R package for analysis of interpopulation sexual dimorphism differences using summary statistics

Area of Science:

• Bioanthropology
• Statistical Genetics
• Computational Biology

Background:

• Sexual dimorphism in traits varies across populations.
• Bioanthropological studies often report sex differences but lack statistical rigor and data sharing.
• Quantifying and comparing the degree of sexual dimorphism across samples is crucial but methodologically challenging.

Purpose of the Study:

• Introduce TestDimorph, the first R package for testing and comparing sexual dimorphism degrees across multiple samples.
• Provide researchers with a robust tool to analyze inter-sample variations in sexual dimorphism using summary statistics.
• Facilitate reproducible bioanthropological research by enabling direct comparison of dimorphism patterns.

Main Methods:

• Implemented univariate and multivariate analytical approaches for comparing sexual dimorphism across two or more samples.
• Utilized established statistical methods including ANOVA, mixture intersection index, dissimilarity index, and Hedges' g with confidence intervals.
• Developed functions for data simulation and extraction of summary statistics directly within the package.

Main Results:

• Demonstrated the application and functionality of TestDimorph using built-in datasets.
• Showcased the package's capability to perform comprehensive analyses of sexual dimorphism from summary statistics.
• Validated the utility of univariate and multivariate methods for inter-sample dimorphism comparisons.

Conclusions:

• TestDimorph offers a comprehensive solution for analyzing and comparing sexual dimorphism across different samples.
• The package enhances bioanthropological research by providing accessible, statistically sound methods for dimorphism analysis.
• Facilitates data sharing and reproducibility by working with summary statistics and offering data simulation capabilities.