THE SCIENCE
Not sampled. Not estimated.
Most people get a standard dose and then everyone waits to see how it goes. Your genes are one reason two people respond to the same drug completely differently — and that is measurable.
Where the guidance comes from
We do not invent our own interpretations. Every gene–drug recommendation in your Medication Check is based on published evidence from four bodies:
Clinical Pharmacogenetics Implementation Consortium
US Food and Drug Administration labelling
Dutch Pharmacogenetics Working Group
Pharmacogenomics Knowledge Base
How your result is expressed
For each gene we report, your inherited variants are combined into a metaboliser status. That status is what changes how a standard dose behaves in your body.
Your report then sorts each medication into standard precautions or use with caution. What to do about it is a conversation for you and your prescriber.
What we look at
The Medication Check covers 50+ clinical-grade medications across anesthesiology, cardiovascular, gastrointestinal, immunology and oncology, infectious disease, pain management and psychiatry — including warfarin, clopidogrel, the statins, the proton pump inhibitors, and many common antidepressants.
Genotyping reads selected positions across your genome, and an imputation step expands that to a much larger set of variants. The imputation method behind our platform benchmarks at 99.9% accuracy against reference data — measurably ahead of the standard tools it was compared with. That is imputation accuracy, not a claim about diagnostic accuracy.
Why whole genome sequencing is a different thing
Genotyping samples your genome at chosen positions. Whole genome sequencing reads all of it — roughly three billion base pairs — on Illumina next-generation sequencing at 30× coverage. Thirty times means every base is read an average of thirty times over. Not sampled. Not estimated.
This matters because some genes are structurally complex and genotyping cannot fully resolve them. Sequencing also surfaces rare clinically significant variants that standard genotyping simply cannot detect. Sequence once, and the data is yours to keep.
Published, not just asserted
The analysis platform behind Check Genetics has been described in six peer-reviewed papers, including in Nature Scientific Reports, Nature Communications, PLOS ONE and BMC Bioinformatics. Its polygenic models were validated against independent datasets covering more than two million individuals, drawn from cohorts including the UK Biobank, FinnGen, BioBank Japan and the Million Veteran Program.
Being straight with you about the limits
A genetic report does not tell you whether you have a condition, and it is not a prescription.
Age, kidney and liver function, other medications and diet all affect how you respond to a drug.
Genetic reference data is drawn largely from people of European ancestry, so results can be less precise for other ancestries. We would rather say so than pretend otherwise.
Never start, stop or change a medication based on a report alone.
Your data
Samples are processed in a CLIA-certified laboratory. The platform holding your results is HIPAA and GDPR compliant, SOC 2 audited, and encrypted with AES-256.
Sharing your results with a practitioner is optional, requires your explicit consent, and can be withdrawn at any time from your portal or by emailing info@checkgenetics.com.
See the testsAll reports are for educational and informational purposes only and are not a substitute for professional medical advice. You must be 18 or over. Testing is not available to residents of New York, New Jersey or Rhode Island.