Abstract: We develop a statistical model and computational algorithm for combining already completed analyses of a number of similar datasets. Individual analyses are assumed to be fit independently using previously written standalone software that fits a computationally intensive Bayesian model using Markov Chain Monte Carlo (MCMC) simulation. Constructing a large complex model involving all the original datasets is time consuming and would require rewriting the existing standalone software. Instead, our strategy is to use the existing MCMC samples from the individual posteriors. We place a hierarchical regression model across the individual analyses for estimating parameters of interest within and across analyses. We use a Mixture of Dirichlet processes (MDP) prior for the parameters of interest to relax parametric assumptions and to ensure the prior distribution for the parameters of interest is continuous. We use an iteratively reweighted importance resampling algorithm within Gibbs to sample values of the individual parameters. We demonstrate our approach on a set of phylogenetic analyses of HIV-1 nucleotide sequence data. These analyses were done independently using the phylogenetic software MrBayes that fits a computationally intensive phylogeny model to aligned nucleotide sequence data using MCMC simulation.