Currentyl little john explicitly skips bams with more than 50K reads. We need to handle this transparently to the user and let them know about the issue.
We could also optionally handle this, although little john is designed to handle small batch streaming data.