The same pharmacogenomics that sits on our condition pages, read from the other end. People rarely arrive asking about SLCO1B1 — they arrive having been prescribed a statin. Each page below collects what published guidelines say about one drug, names the genes involved, and links to the variants your own file would contain.
Supplements — kept apart, with the evidence stated on each →
None of this is dosing guidance. Where a clinical guideline exists it is written for prescribers, and the pages say so. Nothing here is a reason to start, stop or change a medication.
Genotype explains how much drug reaches the blood. It does not explain whether the drug will work, and the guideline is careful about the difference in a way the marketing around this rarely is.
CYP2C19
See what the research says → AntifungalA drug whose blood level is already measured routinely, because it varies so much between people. Genotype explains a good part of why two patients on the same dose land in different places.
CYP2C19
See what the research says → AntiseizureTwo HLA alleles, two different reactions, and recommendations that name populations rather than applying to everybody. The population part is not a footnote on the guidance — it is the guidance.
HLA-B
See what the research says → AntiseizureTwo genes doing two different jobs on one prescription: one decides how much drug there is, the other decides whether it is safe to start at all.
CYP2C9
See what the research says → Urate-loweringThe genotype here is checked before the first dose and never again, and what it predicts is not how well the drug works but whether it is safe to start. It is also the clearest case on this site of a test that only makes sense for some populations.
HLA-B
See what the research says → ImmunosuppressantThis drug's own label carries a warning that runs the opposite direction from most on this site: not a dosing adjustment based on a metabolism gene, but a caution rooted in a separate, rare genetic disease -- the same one behind a severe inherited form of gout.
HPRT1
See what the research says → AntipsychoticOne enzyme decides how much of these a person is exposed to, and the side effects it changes are the ones that make people stop — which in this illness is the largest risk there is.
CYP2D6
See what the research says → AnalgesicThe painkillers people reach for without thinking, and the only reason genotype comes into it is that the harms are dose-dependent and the dose in your blood is not the dose on the packet.
CYP2C9
See what the research says → ADHD (non-stimulant)A genotype here changes the dose, not the diagnosis. Nothing about ADHD is decided by a gene test, and the guideline is about exposure.
CYP2D6
See what the research says → AntiviralThe clearest pharmacogenetic story medicine has had, and it is over. Two genotypes told you whether this regimen would work and whether it would make you anaemic — and then better drugs made the question stop mattering.
IFNL3 · ITPA
See what the research says → Multiple sclerosis therapyOne of the few medicines that cannot be prescribed at all until the test has been done. Here the genotype is not advice about a dose — it is part of the prescription.
CYP2C9
See what the research says → AntiplateletA prodrug: it does nothing until an enzyme switches it on. People whose enzyme works poorly get less of the drug they were prescribed, and this is one of the few places where a guideline says to use a different medicine.
CYP2C19
See what the research says → ChemotherapyFluorouracil and capecitabine, the backbone of colorectal cancer treatment. One enzyme clears most of the dose, and the small number of people who lack it can be harmed severely by an ordinary dose.
DPYD
See what the research says → ChemotherapyThe enzyme that disposes of its active metabolite is the same one behind Gilbert syndrome — a harmless quirk of bilirubin that turns into a dosing problem the moment this drug is prescribed.
UGT1A1
See what the research says → AntiemeticAlmost everywhere else on this site the slow metaboliser is the one at risk. Here it is the fast one: the drug leaves before it has done its job, and the nausea it was given for is not controlled.
CYP2D6
See what the research says → Hormonal therapyTaken for years after early breast cancer, and it too is a prodrug. Whether the enzyme that activates it should change prescribing is one of the longest unsettled arguments in pharmacogenomics.
CYP2D6
See what the research says → AntiplateletA newer alternative to clopidogrel for acute coronary syndromes, taken directly rather than needing an enzyme to switch it on. Genetics moves its blood levels a measurable amount — and, unlike clopidogrel, that movement has not been shown to change how well it works.
SLCO1B1 · CYP3A4/CYP3A5 region
See what the research says → AntihypertensiveThe fast and slow acetylators were described in the 1950s, which makes this one of the oldest known examples of inherited variation in how a body handles a drug — and it took until December 2025 for a dosing table to exist at all.
NAT2
See what the research says → Antihypertensive (thiazide diuretic)One of the most widely prescribed blood pressure drugs, and one already known to raise blood sugar and triglycerides in some people. A common variant moves how much.
NELL1
See what the research says → Beta blockerBlood pressure and pulse are measured at every review anyway, so here the genotype explains a response rather than predicting one — which is a more honest description of most pharmacogenomics than the field usually gives.
CYP2D6
See what the research says →Showing 20 of 55 · page 2 of 3
The same pharmacogenomics already published on our condition pages, read from the drug side instead: what published guidelines say about one medicine, which genes are involved, and the variants your own file would contain.
No. Where a clinical guideline exists it is written for prescribers, and nothing here is a reason to start, stop or change a medication — that belongs with the clinician or pharmacist managing treatment.
Because almost nothing sold as a supplement has a prescribing guideline behind it the way a drug does, and listing the two together would lend the second the authority of the first.
By gene symbol as a whole word, gathered from the pgx_notes already published on the condition pages that discuss each drug — not a second, separate copy of that information.