package:org.seqcode.deepseq

This package aims to provide a set of classes that allow you to conveniently load sequencing experiment data from ReadDB or file sources. 
Experiment relationships are kept in a tree structure that represents conditions, (controlled) replicates, and samples.
Experiment types and targets are also indexed. 

Command-line initialization support is provided by ExptConfig. 
The main handle to all underlying data representations is via ExperimentManager. 

The package supports two options for data caching: 
	1) Load all data into memory (resource hungry for more than a couple of experiments)
	2) Maintain local file caches - load each chromosome into memory one at a time, for example. In this option, each experiment has to be entirely loaded into memory before writing to files, so there is still some memory overhead.

Classes in this package depend on classes in org.seqcode.gse and org.seqcode.genome for interacting with the database and navigating genomes. 
However, care should be taken to avoid depending on other classes (e.g. anything in org.seqcode.projects).  

History:
--------
Nov 18th 2014: Adding paired-end support into HitCache and some HitLoaders.
Aug 3rd 2014: Reorganized to expand the levels of information stored on each experiment (i.e. via ExperimentTarget & ExperimentType), added local file caching support, and made the package independent of the old MultiGPS loading approach. 
May 6th 2014: Package started, adapting from experiment management system in MultiGPS. 

