Odetta is a set of tools for discovering and analyzing novel transcript isoforms using paired-end RNA-Seq data.

External software is used at various points in the pipeline:

• CASHX is used for sequence alignment.
• multisplat is used to discover splice junctions.
• GMB is used to discover novel gene models.

TODO TODO note about rtree

Using Odetta might look like this...

TODO

You can set up a mrjob configuration. For example...

mrjob.conf

runners:
  local:
    base_tmp_dir: /path/to/tmp/dir
      jobconf:
        mapreduce.job.maps: 8
        mapreduce.job.reduces: 7

Use the configuration with...

python example.py --conf-path ./mrjob.conf input_file > output_file

mapreduce.job.maps and mapreduce.job.reduces are particularly useful for utilizing all available processors when running locally (i.e. not Hadoop).

The mrjob docs describe all available options.

Odetta uses mrjob for map/reduce processing. mrjob makes developing and running map/reduce easy, both locally and on Hadoop.

I have not tested this on Hadoop.

nose is used for unit testing. You can run the tests using nosetests tests/.