P2C2M-package: Posterior Predictive Checks of Coalescent Models
In P2C2M: Posterior Predictive Checks of Coalescent Models

P2C2M provides functions to read default output from BEAST (Drummond and Rambaut 2007) and *BEAST (Heled and Drummond 2010) and conduct posterior predictive checks of coalescent models (Reid et al. 2014) with the help of data simulation and summary statistics under various settings.

Installation Instructions

To use P2C2M, the default version of Python must be set to Python 2.7. Users of unix-like operating systems can insure that this requirement is fulfilled by setting the following alias:

echo 'alias python=python2.7' >> ~/.bashrc

Mandatory and optional dependencies of P2C2M can be installed automatically via two installation scripts that are co-supplied with the package. These scripts were designed for unix-like operating systems and are located in folder /exec. To use these installation scripts, a correct configuration of python2-setuptools is required. Users of unix-like operating systems can insure a correct configuration by setting the following alias:

echo 'alias python-config=python2-config' >> ~/.bashrc

To execute the R installer, please run the following commands in R:

source('/path_to_P2C2M/exec/P2C2M.installRlibs.R'); p2c2m.install()

To execute the Python installer, please run the following command in a terminal:

python /path_to_P2C2M/exec/P2C2M.installPylibs.py

Special Note for MacOS

Users of the MacOS operating system need to install the dependencies manually. Prior to their installation, please confirm that file '/usr/bin/python2-config' exists in your file system and that it points to the Python 2.7 executable. Please refer to http://cran.r-project.org/bin/macosx/RMacOSX-FAQ.html on how to install R packages manually. For the manual installation of Python libraries, please refer to http://docs.python.org/2/using/mac.html

Study Design Requirements

In the user-supplied data set, every species should be represented by at least two alleles. Species that are represented by only a single allele, by contrast, must be specified via option "single.allele" and thereby are not included in the calculation of the summary statistic 'GSI'; misspecifications causes P2C2M to print the error message 'Error: given group represents one or fewer taxa. Cannot compute index.').

Input File Requirements

In order to execute P2C2M, a user must provide a directory with three different types of input files: (a) a file that contains species trees, (b) a file that contains gene trees for each gene under study, and (c) an XML-formatted file generated by BEAUTi, the input generator of BEAST (Drummond and Rambaut 2007). A species tree file contains a draw of s generations from the posterior distribution of species trees. Each gene tree file contains an equally large draw from the respective posterior distribution of ultrametric genealogies. Please note that the generations recorded in the species tree file must match those in the gene tree files exactly. The input file generated by BEAUTi is formatted in XML markup language and represents the starting point for a species tree inference in *BEAST. Here, it provides information on allele and species names, the association between alleles and species, and ploidy levels to P2C2M.

File Name Requirements

The following requirements for input file names are in place: The species tree file must be named 'species.trees'. Each gene tree file must be named 'g.trees', where the letter g is substituted with the actual name of the gene. The name of the xml-formatted input file is not constrained and at the discretion of the user. Please be aware that P2C2M uses the name of the xml-formatted input file name to label all subsequent output of the package.

Michael Gruenstaeudl, Noah Reid

Maintainer: Michael Gruenstaeudl gruenstaeudl.1@osu.edu

Drummond, A.J. and Rambaut, A. (2007) BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evolutionary Biology, 7, 214.

Gruenstaeudl, M., Reid, N.M., Wheeler, G.R. and Carstens, B.C., submitted. Posterior Predictive Checks of Coalescent Models: P2C2M, an R package.

Heled, J. and Drummond, A.J. (2010) Bayesian inference of species trees from multilocus data. Molecular Biology And Evolution, 27, 570–580.

Reid, N.M., Brown, J.M., Satler, J.D., Pelletier, T.A., McVay, J.D., Hird, S.M. and Carstens, B.C. (2014) Poor fit to the multi-species coalescent model is widely detectable in empirical data. Systematic Biology, 63, 322–333.

P2C2M documentation built on May 2, 2019, 8:24 a.m.