Follow the links below to download the program, either the source code if you wish to compile on your own machine, or executable files.
| Version | Source, including examples and manual |
Binaries | Manual |
| 1.4.0 (July 12, 2010) | tarball | executable files | |
| 1.3.2 (April 4, 2010) | tarball | ||
| 1.3.1 (10/2009) | tarball | ||
| 1.3.0 (1/2009) | tarball | executable files | |
| 1.2b (12/2008) | tarball | executable files |
For those who wish, you can get the latest version from this git repository.
Here are the slides used for a demo of the program, at the workshop "Estimating Species Tree" (University of Michigan, Jan. 2009).
Version 1.4.0 (June 28, 2010) provides an estimated
population tree. The method used is based on quartet concordance factors
and is consistent under the coalescent model.
Option -sg was added to skip genes, and option --opt-space to
handle very large data sets.
Input files no longer need to use the same root in their tree format.
7/2010: a fix was applied for compilation under a wider range of systems.
Version 1.3.2 (March 8, 2010) can use input files with
different sets of taxa, including for analysis the taxa that are common
to all genes only.
The user can also exclude taxa from the analysis, using a 'pruning' input file.
An important bug was fixed in the sampling probabilities used in the group update.
4/2010: a bug was fixed with mbsum.
Version 1.3.1 (10/2009) can handle any number of taxa. mbsum was improved to be more flexible with line breaks and taxon numbers.
Version 1.3.0 (1/2009) includes the possibility to indicate the names of input files (one input file per gene) into a single file, and the use of translate tables to check that all genes use the same ordering of taxa.
Version 1.2b (12/2008) includes multiple independent runs and a bug fix regarding the group update.