A 2026 NIH study found WGS outperforms DNA arrays on pharmacogenomics accuracy
Most consumer DNA chip tests include a pharmacogenomics section -- a report that estimates whether you are a fast, slow, or normal metabolizer of specific drugs, which affects how well those drugs work at a standard dose. The key question for a kit buyer is whether the underlying technology can reliably detect the variants that actually matter. A July 2026 study published in The Journal of Clinical Pharmacology (Numanagic et al., PMID 42405643) offers the most direct answer to date. Researchers at the National Institutes of Health in Bethesda, Maryland, tested 293 patients using whole-genome sequencing, exome sequencing, and a commercial genotyping array, then cross-checked disagreements against long-read sequencing as an independent reference. Genome sequencing achieved 99% baseline accuracy and resolved 5% of array-based pharmacogenomic calls that the chip had left as unknown or ambiguous. The authors describe it as the first study to demonstrate that genome sequencing provides superior precision and accuracy over an array-based pharmacogenomics platform.
The 5% gap is not a rounding error in a clinical setting. The genes most relevant to pharmacogenomics -- CYP2D6 for antidepressants and codeine, CYP2C19 for clopidogrel and proton pump inhibitors, SLCO1B1 for statins -- include both common variants that chips handle well and rare or structurally complex variants that chips cannot consistently resolve. A chip is designed to look for a finite list of known positions; a genome sequence reads every position and can be reanalyzed as variant databases expand, without retesting. The NIH study found that genome sequencing caught clinically significant inaccuracies, including critical variants missed by the array, that would have gone undetected by chip-only testing.
A second 2026 study, published in the European Journal of Human Genetics (PMID 41639595), approached the question from a different angle. Researchers analyzed whole-genome sequencing data from 1,000 patients who had already been sequenced at a Dutch hospital for rare disease diagnosis. Clinically relevant pharmacogenomic variants appeared in 97% of the cohort. Among the 359 patients with available prescription records, approximately 30% had at least one drug for which their genetic profile suggested a possible therapy change. The finding points to a practical implication for consumer kit buyers: a 30x genome sequenced for one purpose already contains drug-response information that can be extracted without a separate pharmacogenomics test.
Among the consumer WGS providers on this site's roster, several include pharmacogenomics in their standard report. Veritas Genetics provides a clinical pharmacogenomics report alongside its myGenome kit, delivered through a physician-supervised process, per its product page. Nucleus Genomics lists pharmacogenomics as a component of its health report, per its website. Sequencing.com delivers PGx analysis through its app library following sequencing, according to its product documentation. Dante Labs advertises drug-response reporting in its Whole Genome product, per its website, though customer reviews on Trustpilot and similar platforms have consistently noted turnaround delays that affect when any report arrives. None of these consumer reports has been independently benchmarked against the NIH study's methodology. The NIH research establishes that whole-genome sequencing is the correct underlying platform for pharmacogenomics -- what it does not establish is that every consumer report built on top of it is equally useful.
For a buyer weighing a 30x WGS kit against a chip-based test, the 2026 research adds a concrete data point: the chip's pharmacogenomics section will miss a portion of variants the chip was not designed to find, and that portion is not negligible in clinical practice. A genome sequence has no such architectural limit. That advantage only pays off if the report layer actually uses the full variant data -- something worth checking on any provider's sample report before purchasing. Any pharmacogenomic finding that concerns a current medication is worth discussing with a pharmacist or physician before acting on it.