OBJECTIVE: Evaluating the relationships between the shape and biomechanical function of bone cross-sections can contribute novel insights towards human functional and evolutionary morphology. However, this research involves unique analytical and statistical challenges when comparing complex and multidimensional shape data to multivariate biomechanical variables. We developed the CSGM package in the R programming language to streamline statistical hypothesis testing on the geometric properties and shape of bone cross-sections, offering unique interactive and informative visualization plots. This package uses a variety of popular statistical inferential techniques including correlation, covariation, classification, and prediction. By applying a novel nested hypothesis testing approach, users can efficiently analyze complex morphofunctional relationships in parallel. MATERIALS AND METHODS: We present various functions within the CSGM package that can analyze and visualize three-dimensional shape relationships. In addition, we highlight dedicated functions which evaluate pairwise relationships between bone cross-sectional shape and biomechanically relevant variables. The effectiveness of this automated hypothesis testing approach is demonstrated through the use of two associated, complex datasets generated from cross-sections of the mandibular corpus in three modern human collections. RESULTS: The functions of our package helped reveal prominent shape asymmetry in the study sample which also asymmetrically impacts the bending resistances and breaking strength properties of the mandibular corpus. DISCUSSION: The CSGM package offers a series of functions that can test morphofunctional relationships by incorporating a nested hypothesis modeling approach to statistical analysis and interactive graphic visualizations. Thus, CSGM is a useful and powerful analytical toolkit to interpret complex data relationships.
Keeling et al. (Fri,) studied this question.