I tried using subset sequences from a previous bismark run (see Jupyter notebook and lab notebook) to test differences in 1x, 3x, and 5x coverage in methylkit (see lab notebook). These subset sequences did not have enough data.
I have three options:
- Find other deduplicated, sorted, and indexed C. virginica full sample data that used -score_min L,0,-1.2 (maybe from @sr320's runs...?) to test coverage with
- Finish this
bismark run on hummingbird and use that output. The files are currently on gannet
- Wait for my data to eventually come from Mox
I'm leaning towards option 2. Any objections/things I'm not considering?