A shape analysis framework for small animal phenotyping with application to mice with a targeted disruption of Hoxd11

This paper proposes the use of high-dimensional, point-based shape models for the study of phenotype, which offer some important advantages over traditional morphometries, including a more detailed representation of shape and a more systematic approach to shape statistics. Our approach allows for the analysis of shape at both global and local scales and can automatically extract lower- dimensional shape features useful for phenotypic characterization. We also demonstrate how covariates can be included in this framework. We apply the proposed modeling framework to a study of bone shape in mice with a targeted disruption of the Hoxd11 gene.