SVDquartets: Singular Value Decomposition
Scores for Species Quartets
SVDquartets is now implemented in PAUP*!
You can
download the latest version of the software
here: http://paup.phylosolutions.com.
All analysis should be carried out using this new version of the
software. The earlier versions are maintained below for any analyses
that pre-dated the implementation of the method in PAUP*.
Descriptions of the method can be found in the following references:
Chifman, J. and L. Kubatko. 2014. Quartet inference from SNP data
under the coalescent, Bioinformatics, 30(23): 3317-3324.
Chifman, J. and L. Kubatko. 2015. Identifiability of the unrooted
species tree topology under the coalescent model with
time-reversible substitution processes, site-specific rate
variation, and invariable sites, Journal of Theoretical Biology 374: 35-47.
Published analyses that use the method should also cite PAUP*.
Older Versions:
Version 1.0
Laura S. Kubatko and Julia Chifman
Department of Statistics
The Ohio State University
Columbus, OH 43210 lkubatko@stat.osu.edu
Copyright 2014 by Laura Kubatko. This software
is provided "as is" without warranty of any kind. In no event
shall the author be held responsible for any damage resulting from
the use of this software. The program package, including source
codes, executables, and documentation, is distributed free of
charge. This software is covered under the Gnu GPL (General Public
License).
The method is described in the
following papers:
Chifman, J. and L. Kubatko. 2014. Quartet inference from SNP data
under the coalescent, Bioinformatics, 30(23): 3317-3324.
Chifman, J. and L. Kubatko. 2014. Identifiability of the unrooted species tree topology under the coalescent model with time-reversible substitution processes, available on arXiv.
About the Program
SVDquartets is a program to compute a score based on singular value
decomposition of a matrix of site pattern frequencies corresponding
to a split on a phylogenetic tree. These quartet scores can be used
to select the best-supported topology for quartets of taxa, which in
turn can be used to infer the species phylogeny using quartet
methods.
The SVDquartets program relies on
GSL, which must be
installed on your system prior to running SVDquartets. You may have
to change the Makefile to give the correct path to the GSL
libraries.
The SVDquartets program will randomly sample a user-provided number of
quartets from among the taxa provided in the input file called
data.phy, and will estimate the relationship among the quartets.
These inferred relationships are output to the file called
"quartets.out".
The input file must be PHYLIP-formatted. An exception is that the
first row of the file contains not only the number of taxa and the
number of sites, but also a "0" and the number of quartets to be
sampled. See the example input file.
SVDquartets will only generate quartet relationships. You will
need to use a quartet assembly program to build your overall
phylogenetic estimate. We recommend Quartet MaxCut (Snir and Rao,
MPE 62(1): 1-8, 2012).
Instructions:
Compile the programs from the source codes:
(a) Copy "Makefile_scores" to "Makefile" and type "make".
(b) Copy "Makefile_quartets" to "Makefile" and type "make".
Place your data in PHYLIP format in a file called "data.phy" (an example is placed in the file snakes.phy in the downloaded files).
Check to see that the file "splits" (included in the download) is
in the current directory
Issue the command " ./SVDquartets ".
The output file is called "quartets.out". This is the file that
would be used in the quartet assembly programs. It is formatted
correctly for Quartet MaxCut, but may need to be modified for other
programs.
Acknowledgements
Continued development of SVDquartets is based upon work supported by the
National Science Foundation under Grant DMS 1106706.
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
Contact Info
Questions or comments concerning the program can be e-mailed to me at lkubatko@stat.osu.edu. Please also e-mail me if you'd like to be advised when updated versions of the program become available.