Researchers at Rice University report that evolved variants of a compact CRISPR enzyme can bind previously weak PAM configurations, expanding target access for editing genes linked to coronary heart disease. The approach relies on an AI-supported platform, Sequence Display, to generate sequence–activity datasets and rapidly identify protein variants tuned to existing PAM motifs. Using newly evolved SlugCas9 variants, the team edited sites in APOC3 and PCSK9, two genes associated with coronary heart disease risk. The study highlights that target-site editing failures often stem from PAM binding weakness rather than guide choice alone, and that protein engineering can address this constraint more directly. The engineering pipeline is described as enabling candidate identification within about one week, based on rapid dataset creation and deep learning evaluation. The work is experimental in cultured cells, but it frames a platform direction for widening CRISPR coverage for therapeutic and research editing. For the biotech pipeline, the relevance is clear: improved PAM compatibility could reduce the number of “untargetable” sites and strengthen feasibility for gene editing programs where safe locus choice and target accessibility both matter.