To download the alignments, 
-create a work directory: "mkdir work_dir"
-download the alignments (align.tar.gz file) in this directory
-unzip the file "cd work_dir; (g)unzip align.tar.gz"
-extract the alignments: "tar xvf align.tar"

The alignments are organized as follows: 
"work_dir/locus/hum_chp_og/aligned_region/''
where 
``locus'' is the name of one genomic loci comprising a megabase;
``hum_chp_og'' has no particular meaning; 
``aligned_region'' is a directory name describing the aligned region in the locus.

For each alignment, two files are given:
-seq.fasta: the raw sequences of human / chimp / baboon (in this order) which heading describes the origin of the sequence (accession number / start / stop)
-seq.mase: the aligned sequences of human / chimp / baboon in a .mase fromat.

The chimp and baboon sequences were taken from Genbank (Release 133, February 2003). The human sequence was taken from Ensembl (release 8.3).

For more details on the alignments, see the related paper and supplementary material section, or contact Julien Meunier at the following email address: meunier@biomserv.univ-lyon1.fr.