CYP2D6: the false normal that arrives looking like a result
A pipeline that does not treat the CYP2D6 locus specifically does not return an error. It returns *1/*1 — and a poor metaboliser is reported as normal.
CYP2D8P–CYP2D7–CYP2D6 locus at 22q13.2. Above, the same germline BAM a conventional pipeline starts from: reads falling in the near-identical blocks shared with CYP2D7 have no unique destination, receive MAPQ 0 and carry no usable evidence. Below, the same material after a specialised locus-specific analysis. Schematic representation; metrics measured on our own samples.
A whole-genome or exome pipeline that does not treat the CYP2D6 locus specifically does not return an error. It returns *1/*1. A poor metaboliser is reported as normal, with exactly the same appearance of certainty as a real result, and with no alarm anywhere in the chain.
Why this locus, and not another
CYP2D6 sits beside a pseudogene of very high sequence identity. A short read falling in that shared region cannot be placed uniquely: its mapping quality collapses, and what follows in a general-purpose pipeline is a region with no usable evidence in it.
The failure is not that the analysis is wrong. It is that the absence of evidence quietly becomes evidence of normality — and the report goes out saying normal metaboliser, standard dose, for drugs whose CPIC recommendation depends entirely on phenotype: codeine, tramadol, tamoxifen, tricyclic antidepressants.
What we do instead
CYP2D6 is interpreted through a specialised locus-specific analysis rather than inherited from the general-purpose path. In practice that means the locus is resolved — including copy number, whole-gene deletion and duplication — or it is declared not evaluable. There is no third outcome in which it silently looks normal.
Against reference materials
| Material | Diplotype vs. published | Copies | Structure demonstrated |
|---|---|---|---|
| Reference A | concordant | 2 | Loss-of-function allele on a normal background |
| Reference B | concordant | 1 | Whole-gene deletion |
| Reference C | concordant | 3 | Duplication with a reduced-function allele |
The copy numbers are internally coherent with the expected structure — deletion → 1, duplication → 3 — which is what makes them meaningful: they validate the complete chain, not only the final assignment. The materials used and the full results are identified in the technical note. Verification continues with additional reference materials.
The rule that holds the report up
When effective coverage of a gene does not reach threshold, the gene is not reported. The rule is enforced consistently before reporting — it is not a judgement made at the end.
This is a design decision, not a limitation. The alternative to saying not evaluable is not silence: it is asserting normality with no basis. A laboratory can repeat, extend or refer a sample marked as not evaluable. It can do nothing with a false normal it does not know is one.
Declared scope
Resolved: null and reduced-function alleles; whole-gene deletion; duplication and multiplication with copy number; metaboliser phenotype and activity score; explicit not-evaluable declaration by coverage.
Requires confirmation or falls outside scope: which specific allele is duplicated when phase is not resolved; rare hybrid alleles; phenoconversion by strong inhibitors, which is not genotypic and must be assessed against active medication; alleles not described in the current PharmVar nomenclature.
See it run on your data
Request a demo of the full chain — from raw reads to a reviewable, signable report.