Install instructions for hmm script on Ubuntu or Debian systens

1 - install the ghmm package

1.1 Install depending packages (see www.ghmm.org for details)
sudo apt-get install debhelper autotools-dev libglib2.0-0 libgl1-mesa-glx  mime-support libmagic1 locales file libxml2-dev libasound2 lib32asound2 pkg-config automake libpcre3 gawk libtool
1.2 Download ghmm from
http://www.cin.ufpe.br/~igcf/Metabolites/scripts/hmm/ghmm0.8.zip (for 64 bits) or
http://www.cin.ufpe.br/~igcf/Metabolites/scripts/hmm/ghmm.tar.gz (for 32 bits)
1.3 Extract file, go into ghmm directory and run commands
  ./autogen.sh
  ./configure
  make
  sudo make install

2 - Instal Pymix
2.1 Install depending packages (see www.pymix.org for details)
sudo apt-get install python-dev python-numpy libgsl0-dev
2.2 Download pymix from http://www.cin.ufpe.br/~igcf/Metabolites/scripts/hmm/pymix-0.8a.zip
2.3 Extract file, go into pymix directory and run commands
  python setup.py build
  sudo python setup.py install


3 Install GQL
3.1 Install depending packages
 sudo apt-get install python-numpy python-scipy
2.2 Download GQL from
http://www.cin.ufpe.br/~igcf/Metabolites/scripts/hmm/GQL2.1.zip
2.3 Extract file, go into pymix directory and run commands
 python setup.py build
 sudo python setup.py install


4 Download http://www.cin.ufpe.br/~igcf/Metabolites/scripts/hmm/train_rank.py 
and use it to run experiments(see train_rank.py for options)

- For example to run the HMM with 2 states on co2 data (obtained at http://www.cin.ufpe.br/~igcf/Metabolites/data/Biological%20Data%20Python/) run

python train_rank.py -s 2 arab_met_co2.txt arab_co2.txt arab_co2_hmm2

- Results files arab_co2_hmm2.res contains the likelihood of genes (lines) vs. metabolites (collumns).