Annals of Applied Statistics
- Ann. Appl. Stat.
- Volume 7, Number 2 (2013), 989-1009.
Bayesian alignment of similarity shapes
We develop a Bayesian model for the alignment of two point configurations under the full similarity transformations of rotation, translation and scaling. Other work in this area has concentrated on rigid body transformations, where scale information is preserved, motivated by problems involving molecular data; this is known as form analysis. We concentrate on a Bayesian formulation for statistical shape analysis. We generalize the model introduced by Green and Mardia [Biometrika 93 (2006) 235–254] for the pairwise alignment of two unlabeled configurations to full similarity transformations by introducing a scaling factor to the model. The generalization is not straightforward, since the model needs to be reformulated to give good performance when scaling is included. We illustrate our method on the alignment of rat growth profiles and a novel application to the alignment of protein domains. Here, scaling is applied to secondary structure elements when comparing protein folds; additionally, we find that one global scaling factor is not in general sufficient to model these data and, hence, we develop a model in which multiple scale factors can be included to handle different scalings of shape components.
Ann. Appl. Stat., Volume 7, Number 2 (2013), 989-1009.
First available in Project Euclid: 27 June 2013
Permanent link to this document
Digital Object Identifier
Mathematical Reviews number (MathSciNet)
Zentralblatt MATH identifier
Mardia, Kanti V.; Fallaize, Christopher J.; Barber, Stuart; Jackson, Richard M.; Theobald, Douglas L. Bayesian alignment of similarity shapes. Ann. Appl. Stat. 7 (2013), no. 2, 989--1009. doi:10.1214/12-AOAS615. https://projecteuclid.org/euclid.aoas/1372338476
- Supplementary material: Simulation methods and a normal approximation for the halfnormal-gamma distribution. We describe an acceptance-rejection method for simulating from the halfnormal-gamma distribution and investigate its efficiency over a range of parameter settings. We also investigate further the normal approximation to the halfnormal-gamma distribution, which we use to obtain efficient proposals in our Metropolis updates. We show that the approximation is best for parameter values where the acceptance-rejection method is less efficient, and hence that the two simulation methods complement each other well.