Version 1.2.0

Fixes bugs introduced in v1.1 and switches back to v1.0 strategy for estimating gene content.
Code and reference database should be stable going forward.

1. uses new database format
   -pan-genome genes defined at 99% identity threshold (same as v1.0)
   -includes mapping of genes to gene clusters at lower % identity thresholds (75,80,85,90,95)
   -enables sensitive read mapping while giving flexibility in defining pan genome gene families
   for other info, see: http://lighthouse.ucsf.edu/MIDAS/README.txt

2. updated build_midas_db.py
   -provides same features described in (1)

3. disabled map quality filtering in pan-genome pipeline (--mapq)
   -when genes are clustered at 99% identity, reads will often map to >1 gene, receive a low mapping quality, and get discarded
   -this feature was disabled to retain these reads

4. bug fixes
   -fixed estimation of marker gene coverage in 'run_midas.py species' (bug introduced in v1.1)
   -fixed normalization by marker gene coverage in 'run_midas.py genes'

5. 'run_midas.py genes' and 'merge_midas.py genes' report # of reads mapped to each gene cluster