/* RAxML-VI-HPC (version 2.2) a program for sequential and parallel estimation of phylogenetic trees * Copyright August 2006 by Alexandros Stamatakis * * Partially derived from * fastDNAml, a program for estimation of phylogenetic trees from sequences by Gary J. Olsen * * and * * Programs of the PHYLIP package by Joe Felsenstein. * * This program is free software; you may redistribute it and/or modify its * under the terms of the GNU General Public License as published by the Free * Software Foundation; either version 2 of the License, or (at your option) * any later version. * * This program is distributed in the hope that it will be useful, but * WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY * or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * for more details. * * * For any other enquiries send an Email to Alexandros Stamatakis * Alexandros.Stamatakis@epfl.ch * * When publishing work that is based on the results from RAxML-VI-HPC please cite: * * Alexandros Stamatakis:"RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models". * Bioinformatics 2006; doi: 10.1093/bioinformatics/btl446 */ December 12 2008 qmmRAxML: Q-matrix mixture RAxML The qmmraxml is based on RAxML-VI-HPC (version 2.2). It implements a class-frequency mixture model for ML inference of protein phylogeny. For the moment, only protein sequences can be used for qmmRAxML. Citation: Huai-Chun Wang, Karen Li, Edward Susko and Andrew J. Roger: A class frequency mixture model that adjusts for site-specific amino acid frequencies and imporves inference of protein phylogeny. BMC Evolutionary Biology (in press). To compile the program, type: make To run the program with the amino acid class frequency mixture model, type: (for tree search starting from a random tree) ./qmmraxmlHPC -m PROTGAMMAJTTF -s cox2.seq -n cox2.out -@ (for tree search starting from a user input tree) ./qmmraxmlHPC -m PROTGAMMAJTTF -s cox2.seq -t cox2.tre -n cox2.out -@ (for fixed tree) ./qmmraxmlHPC -m PROTGAMMAJTTF -s cox2.seq -t cox2.tre -f e -n cox2.out -@ To run the program with a user-defined rate matrix or user-defined amino acid class frequency models, say, myModels.dat, type: (for tree search starting from a random tree) ./qmmraxmlHPC -m PROTGAMMAJTTF -s cox2.seq -n cox2.out -u myModels.dat (for tree search starting from a user input tree) ./qmmraxmlHPC -m PROTGAMMAJTTF -s cox2.seq -t cox2.tre -n cox2.out -u myModels.dat (for fixed tree) ./qmmraxmlHPC -m PROTGAMMAJTTF -s cox2.seq -t cox2.tre -f e -n cox2.out -u myModels.dat Please refer to RAxML-VI-HPC (version 2.2.3) User Manual (http://icwww.epfl.ch/~stamatak/index-Dateien/software/RAxML-Manual.2.2.3.pdf) for details about the command line arguments. -m specifies protein model. -s specifies input sequence file name. -n specifies output file name. -t specifies starting tree. -f e optimizes model+branch lengths for a given input tree (topology is fixed during optimization) under GTRGAMMA only. -@ allows doing amino acid class frequency mixture model. Four class frequency vectors will be conbined with the proetin model (specified by the -m option) and the amino acid frequency vector of the whole dataset (specified by the F component of the -m option), as discussed in the Wang, Li, Susko and Roger's paper cited above. -u specifies user-defined amnio acid substitution matrices and amino acid frequency vectors. For instance, a new rate matrix, LG, was recently reported. If you want to use LG with amino acid class frequency models, you can use myModels.dat, which is included with the release. For your own models, you should modify myModels.dat but keep the same format. Note: the -@ and -u options are specific to qmmraxmlHPC. It is recommendated that the two options should not be used simultaneously. If without -@ nor -u option, the qmmraxmlHPC simply runs like Alexandros Stamatakis's original raxmlHPC (a single rate matrix model). An example sequence file (cox2.seq) and a tree file (cox2.tre) are included with the release. If you have a problem with running qmmraxmlHPC, please contact Huaichun Wang (hcwang@mathstat.dal.ca).