Characterization of NAT2 Using Long-Read Sequencing: Allele, Diplotype, and Phenotype Call Accuracy Compared to Other Testing Strategies.

The NAT2 pharmacogene is essential in drug metabolism, particularly for aromatic amines and hydrazines. Genetic variations in NAT2 categorize individuals as rapid, intermediate, or poor metabolizers based on their acetylation capacity to inform dosing guidelines. Genotyping and short-read sequencing allele calling methodologies are unable to phase variants, leading to ambiguous diplotype and phenotype calls while long-read sequencing technologies not only provide variant detection but also variant phasing to unequivocally establish diplotype. This study utilized long-read sequencing to analyze NAT2 genetic variation in a large cohort of 1828 long-read and 662 paired short-read patient samples from the Genomic Answers for Kids program, comparing call accuracy with short-reads and simulated SNP panel tests. A custom tool, staR-NAT2, identified 11 new star alleles and four new suballeles. Long-reads demonstrated high diplotype accuracy (> 99%) with pb-StarPhase and Aldy, excluding novel haplotypes, while short-reads showed a significant reduction to 64% in diplotype concordance, but only affecting 4% of phenotype calls. Despite poor diplotype resolution and accuracy compared with long-read data, short-read sequencing and a 4-SNP panel accurately predicted phenotype for over 95% of subjects. However, the 4-SNP and 5-SNP panels both resulted in discordant or ambiguous phenotypes for non-white individuals at 4.5% and 25%, respectively. Additionally, population-specific alleles were identified among 1144 unrelated individuals, including NAT2*14 (8.9%) and *43 (3.57%) in Blacks and *7 (7.35%) in Hispanics. These findings underscore the limitations of short-read sequencing and support long-read sequencing as a robust approach for accurate and equitable NAT2 pharmacogenetic testing to guide individualized drug therapy.