What the pharmacogenomics report in your WGS kit actually tells you

Most consumer whole genome sequencing kits now include a pharmacogenomics report as part of the package. The pitch is intuitive: your genes influence how your body breaks down certain medications, and knowing your variants could help you and a clinician choose drugs or doses more precisely. That promise is real, but it does not apply uniformly across the hundreds of medications that appear in vendor marketing materials. Understanding what the report actually covers -- and what it does not -- is worth knowing before you order.

The standard for gene-drug evidence in clinical pharmacogenomics is the Clinical Pharmacogenomics Implementation Consortium, known as CPIC, a shared project of PharmGKB and the Pharmacogenomics Research Network that publishes freely accessible prescribing guidelines at cpicpgx.org. CPIC grades each gene-drug pair by evidence strength and publishes prescribing recommendations only where clinical data is strong enough to justify changing a dose or swapping a drug. A 2019 paper summarizing the CPIC guideline development process counted 19 genes and 46 drugs with published guidelines at that time; the list has grown since. The highest-confidence pairs include a short and recognizable set: CYP2C19 and clopidogrel, the antiplatelet drug; CYP2D6 and codeine; DPYD and fluoropyrimidine chemotherapy drugs such as 5-fluorouracil; TPMT and thiopurines such as azathioprine; and HLA-B*57:01 and abacavir, an HIV medication. A patient who carries two nonfunctional CYP2C19 copies metabolizes clopidogrel poorly -- a finding with well-documented consequences after stent placement and one that can change a clinician's prescribing decision in a meaningful way.

One area where WGS has a concrete advantage over standard genotyping arrays is in reading CYP2D6, a gene involved in metabolizing a significant share of psychiatric drugs, pain medications, and cardiac drugs. CYP2D6 contains copy-number variants and structural rearrangements that chip-based tests routinely miss or call incorrectly. A WGS run reads all positions at depth and can resolve CYP2D6 duplications and rare alleles that arrays cannot. SelfDecode's WGS product page says the test "accurately resolves complex structural variants in drug metabolism genes (e.g. CYP2D6)," per the company's documentation. A 2026 NIH-funded study reported on this site found that WGS also outperforms arrays on pharmacogenomics accuracy more broadly, calling a higher rate of clinically relevant variants correctly.

Where consumer reports diverge is in how far they extend beyond the CPIC core. SelfDecode's product FAQ describes an Expanded PGx report covering approximately 800 medications, per the company's product page. Veritas Genetics describes pharmacogenomics as a reporting category in its myGenome product, per the company's documentation, though the specific gene list and drug count are not itemized in publicly available materials. For a consumer, the number to interrogate is not the total medication count but how many of those entries are backed by Tier 1 CPIC evidence versus lower-confidence associations that are real but not yet clinically actionable. A report returning a weak association note for 795 of 800 drugs is a different product from one returning five high-confidence CPIC results. Vendor materials do not always make that distinction visible.

The FDA authorized the first direct-to-consumer pharmacogenetic test in 2018, for 23andMe, under a de novo classification. In its press announcement, the agency noted that the test "does not describe an association between the detected variants and any specific drug nor whether a person will or will not respond to a particular drug," and required labeling stating that consumers should not stop or change any medication based solely on the results. No pharmacogenomics report from the current roster of consumer WGS providers carries a comparable FDA marketing authorization. That does not make the reports useless; it shifts interpretive responsibility to the clinician reviewing the results. A few things worth checking before ordering: whether the sample report distinguishes between CPIC-graded findings and association-level data, whether findings are tied to a specific clinical recommendation or just a gene annotation, and whether the report recommends clinician review before acting. Those features are easy to verify from a sample report and tell you considerably more about what you are actually buying than the total medication count in the marketing copy.