Are you referring to string-overlap graphs for de novo assembly? In that case, isn't CRISPR addressing another problem?