A Computational Approach for Pharmacogenotype Extraction Using Next-Generation Sequencing and Somatic Whole Exome Sequencing in Precision Oncology

Next-generation sequencing (NGS), including whole exome sequencing (WES) and whole genome sequencing (WGS), is increasingly available to support clinical pharmacogenomics (PGx). This study evaluated the feasibility of repurposing clinical next-generation sequencing (NGS) data, including whole exome sequencing (WES) and whole genome sequencing (WGS), to evaluate pharmacogenomic (PGx) genotypes. The sequencing data was evaluated using the Aldy computational tool. This studies analyzed three distinct patient cohorts based on sequencing data sources: (1) A cohort of 164 patients analyzed using germline WES data and Aldy v3.3, (2) A cohort of 264 patients with germline WES and WGS data analyzed using Aldy v4.4, and (3) A cohort of 719 patients from an institutional molecular tumor board, where both germline and somatic sequencing data were evaluated for PGx genotype concordance.

Concordance rates between germline WES sequencing and clinically determined PGx genotypes were evaluated in the first study, revealing overall concordance of >99% for most pharmacogenes; the exception was CYP2D6, which exhibited lower concordance due to its complex genomic structure. In the second study, Aldy achieved >99% accuracy in diplotype calls using WGS sequencing across the pharmacogenes and identified additional star alleles beyond those detected by targeted clinical genotyping. In the third study, the PGx called alleles from the somatic sequencing were largely concordant with the germline sequencing, although there were several discordances in some genes. Threshold adjustment analysis in the WES cohort revealed that intermediate thresholds (around 20) minimized discordances. Read depth analysis indicated higher read depths in somatic sequencing compared to germline sequencing, though discordances persisted in complex genes like CYP2D6.

These studies collectively demonstrate Aldy’s ability to accurately call diplotypes from NGS data, supporting the repurposing of WES and WGS for PGx in clinical settings. Importantly, analysis of somatic sequencing data indicated that while read depths were generally higher for somatic samples, discordances persisted in complex genes such as CYP2D6, reinforcing the need for cautious interpretation of somatic-derived PGx genotypes.