Amplicon-based long-read sequencing for accurate CYP2D6 gene deletion and duplication detection using CYP2D7 as a reference gene.

Homology between CYP2D6 and its pseudogenes CYP2D7 and CYP2D8 makes copy-number calling difficult; long amplicons across the locus allow CYP2D6 read depth to be referenced against CYP2D7 and copy number to be inferred from it.

The authors describe an amplicon-based long-read sequencing approach (Oxford Nanopore) that detects CYP2D6 single-nucleotide variants and copy-number variations in a single reaction, while comprehensively characterizing the CYP2D7 and CYP2D8 pseudogenes. The entire locus (about 30.2 kb) and then the three homologous genes separately (6.3-6.6 kb) were sequenced. The analysis identified 21 unique CYP2D7 haplotypes and 17 CYP2D8 haplotypes; 57% of the variants detected in the pseudogenes carried nucleotides corresponding to the CYP2D6 reference, and 36 variants found in the pseudogenes are also annotated as functional variants in CYP2D6. The combined CYP2D6/CYP2D7 read-count ratio and reference-to-alternative allele frequency ratio accurately predicted copy-number status, an approach the authors describe as simple and rapid, hence potentially suitable for clinical pharmacogenomics.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

Using a neighbouring pseudogene as an internal reference is an elegant idea: it frees copy-number calling from any external normalization, which is exactly where current CYP2D6 assays fail. The work nonetheless remains an analytical validation: no patient numbers are given, no quantified concordance against a reference method, no diplotype reported under clinical conditions. Before talking about implementation, those figures are needed on a patient series, particularly for duplications of non-functional alleles, the case where a genotyping error changes the dose.

Analysis by Dr Thibaut Benquey