Yeast-test:
Add LPJS scripts

Enhancements:
Improved error output
Improve code readability
fold-change: Add --debug flag
Add R normalization scripts for comparison
Add DESeq2 script for comparison
Replace %Agr with log2(FC): More generally useful

New scripts:
Add fasda filter for filtering fold-change output

Fixes:
fold-change: Add --near-exact to usage message and man page
Bring man pages up-to-date
normalize: Fix MRN bug (divide, not multiply)

abundance:
Update stringtie command to also produce gene-level abundances
Output 0 for effective length for now to clearly indicate that it's invalid
normalize: Add --debug option
Updates for latest libxtend API
Test -> Yeast-test
Add Mouse-test suite with multiple mouse data options
Streamline test scripts

Add --debug
Replace --map-gene with --feature-type for more flexibility
Use extended REs for gff3 feature selection (default = RNA$|transcript$|gene_segment$, same as transcript in GTF)
Add --output-dir
Support |-separated ID aliases (report last alias)
Support outsourcing abundance calculation (default = stringtie)
Several bug fixes and minor enhancements

Updates for libxtend API changes
Improvements to test scripts
Add more diagnostic scripts
Clean up code

Bring fasda abundance to a working state for hisat2 data
Vastly improved test scripts
Various minor enhancements and bug fixes

Verified reasonable results from kallisto output
Fold-changes very close to what Sleuth produced for the same data
P-values largely in the same ballpark as Sleuth's
Improved documentation
Numerous minor fixes

Able to normalize using Mean Rations Normalization and compute fold-change with p-values from a Mann-Whitney U-test.
Output needs plenty of testing and verification, but looks reasonable.