v1.4.0

Explore scaleup to runs with 4-8M reference sequences, including:

• Allow use of precomputed bowtie-build directory
• Increased time for ubr_merge
• New ubr_check_run.py allows monitoring of error and output files from cluster runs.
• New ubr_check_stats.py prints out number of reads at each stage of an ubr_run, and is run automatically at end of the run.
• Initial exploration of cmuts pipeline and i/o with HDF5 files instead of .txt.gz files.