Characterization of complex structural variation in the <i>CYP2D6-CYP2D7-CYP2D8</i> gene loci using single-molecule long-read sequencing

Complex regions in the human genome such as repeat motifs, pseudogenes and structural (SVs) and copy number variations (CNVs) present ongoing challenges to accurate genetic analysis, particularly for short-read Next-Generation-Sequencing (NGS) technologies. One such region is the highly polymorphic <i>CYP2D</i> loci, containing <i>CYP2D6,</i> a clinically relevant pharmacogene contributing to the metabolism of >20% of common drugs, and two highly similar pseudogenes, <i>CYP2D7</i> and <i>CYP2D8</i>. Multiple complex SVs, including <i>CYP2D6/CYP2D7</i>-derived hybrid genes are known to occur in different configurations and frequencies across populations and are difficult to detect and characterize accurately. This can lead to incorrect enzyme activity assignment and impact drug dosing recommendations, often disproportionally affecting underrepresented populations. To improve <i>CYP2D6</i> genotyping accuracy, we developed a PCR-free CRISPR-Cas9 based enrichment method for targeted long-read sequencing that fully characterizes the entire <i>CYP2D6-CYP2D7-CYP2D8</i> loci. Clinically relevant sample types, including blood, saliva, and liver tissue were sequenced, generating high coverage sets of continuous single molecule reads spanning the entire targeted region of up to 52 kb, regardless of SV present (<i>n</i> = 9). This allowed for fully phased dissection of the entire loci structure, including breakpoints, to accurately resolve complex <i>CYP2D6</i> diplotypes with a single assay. Additionally, we identified three novel <i>CYP2D6</i> suballeles, and fully characterized 17 <i>CYP2D7</i> and 18 <i>CYP2D8</i> unique haplotypes. This method for <i>CYP2D6</i> genotyping has the potential to significantly improve accurate clinical phenotyping to inform drug therapy and can be adapted to overcome testing limitations of other clinically challenging genomic regions.