Read Me

You can download a tar ball of GLEAN at:

  http://sourceforge.net/projects/glean-gene

There is no installation procedure (apart from unpacking).  You need
to have R installed, and a handful of Perl libraries:

    BioPerl (preferably, a CVS checkout of bioperl-live)
    List::Util
    FindBin
    YAML
    Storable
    Getopt::Long
    Algorithm::Diff
    Data::Dumper
    Module::Pluggable
    GraphViz (only for some advanced uses)

You invoke glean/bin/glean-lca with a "param.yaml" parameter file in
YAML format (you can find one in data/param-example.yaml) against a
MySQL Bio::DB::GFF database:

  % glean-lca --param myparams.yaml --database fly --user me --password youbetchya > fly.dat

This collates all the data from the database and calculates
performance estimates for each evidence source.  Then you generate
gene models with:

  % cat fly.dat | glean-dp --gff --minlen 45 > fly.gff

--minlen specifies the smallest allowed gene length, in nucleotides;
--gff triggers GFF2 output format (default is the funny format
accepted by GBrowse for user-supplied annotation features).

I'm sure you'll have questions; let me know what breaks first.

Aaron Mackey
ajmackey@gmail.com