Stop filtering short reads in validate_input step. This reads will most likely be remove further downstream and they were not consistently removed from paired-end fastqs (see discussion in the ticket).
Make count more reliable instead of relying on a parallel data structure
Ticket: CH-18285
validate_inputstep. This reads will most likely be remove further downstream and they were not consistently removed from paired-end fastqs (see discussion in the ticket).