Assuming this was true and well documented, this sounds like an interesting classification problem to identify and quickly flag such instances in genome sequencing. As always - this requires good clean data!