Please add a note that input files that contain samples with 0 reads will cause the output count file to be erroneous.