AutoParts is a program that implements Bayesian estimation of phylogeny from partitioned nucleotide sequence data using Markov chain Monte Carlo methods under a Dirichlet process prior model. The evolutionary process may vary across an alignment of nucleotide sequences; for example, between different gene/genome regions, stem vs. loop regions of ribosomal DNA, etc. Conventionally, a series of fixed data-partition schemes are iteratively evaluated by a series of independent analyses to find the 'mixed-model' that adequately captures variation in the evolutionary process. By contrast, AutoParts automates the analysis of partitioned data by treating the data partitioning scheme as a random variable. Under this approach, the prior probabilities for the number of data partitions and the assignment of nucleotide sites to those data partitions are specified by a Dirichlet process prior model, such that the estimate of phylogeny is evaluated by integrating over all possible data partition schemes in the course of a single analysis.
To use Autoparts, load the module:
module load autoparts
This package has the following support level : Supported