in fragment mode, when --estimate_np_by_length enabled, sum up all sampled insert lengths and consider it as the length of polymerase read.
break up the polymerase read length by assuming that first bp is in insert (not in adapter), i.e. the the number of passes is calculated based on length of insert (template) and adapter except for the first (maybe full or partial) pass.
--estimate_np_by_length
enabled, sum up all sampled insert lengths and consider it as the length of polymerase read.