Tetranscripts

Changes
* Added the ability to accept pre-built TE (or gene) GTF indices. This will save time when the same index is used by multiple runs. Pre-built TE GTF indices for human and mouse are available [here](http://labshare.cshl.edu/shares/mhammelllab/www-data/TEtranscripts/Prebuilt_indices/). Please decompress (gunzip) before use.
* Many thanks to Talitha (talithaforcier) for the implementation, and lexa for the suggestion.

Minor changes
* R package loading messages now suppressed when running TEtranscripts.
* Minor update on help message for `--stranded` parameter.
* Fixed errors introduced during repository merge.

Changes
* `--stranded` now defaults to no
* Improved explanation of `--stranded` option in README and help message

Rebranding
* Tool package has been renamed to TEtranscripts.

Minor changes
* Fixes to R code for differential analysis
* Comment lines now handled properly in GTF file (issue 36)

Minor fixes
* Changed DESeq2 `results()` parameter from `alpha=1.0` to `independentFiltering=F` to prevent crashing when using newer versions of DESeq2

Release notes
New tool
**TEcount**
- This is a standalone tool that performs the quantification (& EM) step of TEtranscripts on a single RNAseq library.
- Allows quantification of multiple libraries in parallel that can be later combined for differential analysis. (Issues 5, 17)
- Credit to Devon Ryan (dpryan79) for the idea.

Updated requirements
1. Now requires **pysam 0.9.0 or newer** to perform BAM sorting. (Issues 10, 11)
2. Now implements **DESeq2** for differential analysis (Issue 24)
* DESeq v1.x remains an option for TEtranscripts, but requires the `--DESeq` parameter

Minor fixes
* Changed default number of maximum iterations to 100

Release Notes
1. making "multi" mode as default.
2. make default number of iteration as 10.
3. Fixed the error when the read name contains '/1, /2' for PE BAM files.