mptp

Implementation of a fast species delimitation method, based on PTP (Zhang et al. 2013) with a 64-bit multi-threaded design that handles very large datasets. The tool mPTP can handle very large biodiversity datasets. It implements a fast method to compute the ML delimitation from an inferred phylogenetic tree of the samples. Using MCMC, it also computes the support values for each clade, which can be used to assess the confidence of the ML delimitation. ML delimitation mPTP implements two flavours of the point-estimate solution. First, it implements the original method from (Zhang et al. 2013) where all within-species processes are modelled with a single exponential distribution. mPTP uses a dynamic programming implementation which estimates the ML delimitation faster and more accurately than the original PTP. The dynamic programming implementation has similar properties as (Gulek et al. 2010). See the wiki for more information. The second method assumes a distinct exponential distribution for the branching events of each of the delimited species allowing it to fit to a wider range of empirical datasets. MCMC method mPTP generates support values for each clades. They represent the ratio of the number of samples for which a particular node was in the between-species process, to the total number of samples.