Division of Biostatistics, Washington University School of Medicine, St. Louis, Missouri 63108, USA. todorov@wupsych1.wustl.edu
We provide a general framework for the development of model-free methods for the linkage analysis of multivariate phenotypic data. It is possible within this framework to test both for linkage of a set of phenotypes to one or more markers and for the presence of structural relations among the phenotypes themselves. This report presents the general model, paying special attention to the assumptions that enter its formulation, and outlines the estimation procedures that may be used.