I have changed the way this runs cutadqpt. Before the pipeline did not do well with mutiple gene regions that are partially overlapping and both have potential readthrough. Yes, this is a rare case, but it does happen. I've now reordered the cutadapt process, cutting 5' primers and demultiplexing first, then cutting 3' primers second, if necessary