THEORY

Background

The beast software package is an ambitious attempt to provide a general framework for parameter estimation and hypothesis testing of evolutionary models from molecular sequence data. beast is a Bayesian statistical framework and thus provides a role for prior knowledge in combination with the information provided by the data. Bayesian Markov chain Monte Carlo (MCMC) has already been enthusiastically embraced as the state-of-the-art method for phylogenetic reconstruction, largely driven by the rapid and widespread adoption of mrbayes (Huelsenbeck & Ronquist, 2001) (see Chapter 7). This enthusiasm can be attributed to a number of factors. First, Bayesian methods allow the relatively straightforward implementation of extremely complex evolutionary models. Second, there is an often erroneous perception that Bayesian estimation is “faster” than heuristic optimization based on the maximum likelihood criterion.

beast can be compared to a number of other software packages with similar goals, such as mrbayes (Huelsenbeck & Ronquist, 2001), which currently focuses on phylogenetic inference and lamarc (Kuhner, 2006) (discussed in the next chapter) and batwing (Wilson et al., 2003), which focus predominantly on coalescent-based population genetics. Like these software packages, the core algorithm implemented in beast is Metropolis–Hastings MCMC (Metropolis et al., 1953; Hastings, 1970). MCMC is a stochastic algorithm that produces sample-based estimates of a target distribution of choice. For our purposes the target distribution is the posterior distribution of a set of evolutionary parameters given an alignment of molecular sequences.