Spatial matching of M configurations of points with a bioinformatics application

In this paper, we present a model to deal with the problem of matching M objects or configurations of points. This is a generalization of the model proposed by Green and Mardia (2006). We consider, as a direct and simple application, the case of three configurations with labelled and with unlabelled points. In both cases, we consider data from a microarray experiment of gorilla, bonobo and human cultured fibroblasts published by Karaman et al. (2003). We find out the matchings and the best affine transformation between the projections of genes in a two dimensional space, obtained by a Multidimensional Scaling technique.