I may be missing something, but I think the natural Bayesian analogue just is a model...
On the theoretical side, Szpiro, Rice and Lumley 2010 might be interesting.