Ubuntu instructions

This tutorial was written using Ubuntu Linux.

You will need about 4GB of disk space to keep the hapmap data. The main bulk of data, about 3.5GB, will reside in /var directory.

You can check your current disk usage by typing:

df -h

Downloading the data

cd ~/Desktop
mkdir -p hapmap
cd hapmap
wget http://downloads.sourceforge.net/genedc/genepi-2007-06-20.sql.bz2
wget http://downloads.sourceforge.net/genedc/legend-2006-07.csv.bz2
bunzip2 legend-2006-07.csv.bz2

Don’t uncompress genepi-2007-06-20.sql.bz2 (it’s a quite large file, and you don’t need in uncompressed anyway.)

Preparing database with Hapmap.org data

Please note that these instructions assume you’re working on your local machine. If you want to make an insitution-wide PostgreSQL installation, ask your network administrator how to secure it.

sudo apt-get install postgresql-8.2
sudo -u postgres createuser -s ${USER}
createdb genepi -E utf-8
bzcat genepi-2007-06-20.sql.bz2 | psql genepi

You might see some messages like this, they are harmless.

ERROR:  role "maciej" does not exist

The last command may take up to few hours to complete. Time of completion depends on your computer’s hard disk speed and processing power. Leaving your computer alone until it finishes, is suggested. This is one-time operation. The data will remain in the computer unless you decide to import a new release of Hapmap data. Converting your data to hapmixmap format

Installing required libraries.

sudo apt-get install build-essential libboost-dev flex bison \
libboost-filesystem-dev libboost-program-options-dev \
python-psycopg

Downloading and compiling genedc.

Download genedc tar archive
Compile it with usual ./configure and make

tar xfz genedc-X.Y.Z.tar.gz
cd genedc-X.Y.Z
./configure
make
./genedc --help

To convert your genotypes into hapmixmap format, type:

./genedc -l ../legend-2006-07.csv -i your-file

It will take about 20 seconds.

Preparing training data

From the genedc-X.Y.Z directory:

src/hapmap/cgs-hapmap -g user_genotypes.txt -o training

It will take about 30 seconds to complete the first time. Following runs will take about 5 seconds.