Bug fix and couple minor improvements

[tprodanov]

• Bug fix in output_file_writer.py: in some cases read1 and read2 are identical in output fastq files (and different in output BAM file).
• Skip chromosomes if they are not in input regions and -to 0 or --discard-offtarget.
• ref_func.py: in read_ref if n_handling is random - slightly speed up sequence randomization.

[joshfactorial]

We're encouraging everyone to check out the new, up-to-date repo on github.com/ncsa/NEAT. This repo is no longer actively maintained. Thank you!