separate computing the sample/lane_id names from calculating read counts