jit-able bootstrapping of reads

[jwschroeder3]

I actually made quite a few changes, which I hope won't mess things up too much.

You'll need to install numba, pandas, and dfply for the code to run.

Note the arguments:

• I added a required argument, sample_name_luts, which accepts file names with your sample information from the UofM sequencing core. The script creates a lookup table to associate sample_ids with sample descriptions as you gave them to UofM.

• I changed "ext" to "ChIP" in some arguments.

• I added the "numba" argument. You must include this argument to use the jitted bootstrapping function. If you don't include --numba at runtime, you'll just use the original sampling function.