README

rnaseq 0.1
Copyright 2010 Frederick Ross <madhadron at gmail dot com>

rnaseq is a set of tools for doing statistical inference on RNASeq
data, developed in the Bioinformatics and Biostatistics Core Facility
of the EPFL.  It handles one way linear models, multiply mapped reads,
and scales up to large collections of transcripts.

rnaseq is released under the GNU General Public License 3.0.  A copy
of this license is in the LICENSE file.

Installation
------------

rnaseq has a number of dependencies.  You must have numpy, pyMC,
pysam, sqlite3, NetworkX, and Cython on your system.  You will
probably need to change the 'numpy_include_dirs' variable in setup.py
to point to the numpy headers for Cython on your system.  Then in the
rnaseq directory, run

$ python setup.py build_ext
$ python setup.py build
$ sudo python setup.py install

Known issues
------------

None.

Acknowledgements
----------------

The development of bein was supported by The Bioinformatics and
Biostatistics Core Facility of the EPFL (http://bbcf.epfl.ch/) and
SyBit (http://www.systemsx.ch/projects/systemsxch-projects/sybit/).

Thanks to Jacques Rougemont and Jose Ignacio Molina Clemente for
useful discussions and ideas.

Files
-----

bin/  -- scripts the user runs
  simple_inference.py -- run a one way linear model on two sets of SAM/BAM files

rnaseq/  -- package containing all the working guts of rnaseq.
  __init__.py    -- Construct the public interface of the rnaseq package
  model.pyx      -- Cython source for doing the inference
  subproblems.py -- Find subsets of the transcripts which may be run separately
  bag.py         -- Implementation of a bag data structure for use in multiread mapping
  load.py        -- Functions to assemble SAM/BAM files into a database