The positions here were found by asking three different questions — who develops furred arteries, who has a heart attack early, and who has a heart attack given furred arteries. The answers were not the same genes, which is the most useful thing this page has to say.
Solid lines are connections this site curates. Dashed lines mean the two ends share a research paper — worth knowing, and not a claim that one explains the other.
Coronary artery disease is the narrowing of the vessels that feed the heart, and a heart attack is what happens when one of them closes. Those are two events, not one, and the genetics has been able to tell them apart — which is unusual and worth the page on its own.
A 2011 study did something clever: instead of comparing people with heart disease against everyone else, it used coronary angiograms to ask two separate questions. Who develops coronary atherosclerosis at all (12,393 with it against 7,383 without)? And among people who already have it, who goes on to have a heart attack (5,783 against 3,644)?
The answers differed. rs1994016, near ADAMTS7, came out of the first question — the disease of the artery wall itself. ADAMTS7 encodes an enzyme that acts on the matrix around vascular smooth muscle cells, which is the tissue that thickens as a plaque grows.
The other two came from studies of heart attack. rs12526453 (PHACTR1) and rs1746048 (near CXCL12) were both identified in a genome-wide study of early-onset myocardial infarction — 2,967 cases against 3,075 controls, with replication in up to 19,492 more — where nine loci reached significance and these were among them.
rs6725887 at 2q33, in WDR12, comes from the scan described above — it was the third of the three loci that study found for the first time, alongside PHACTR1 at 6p24 and a region at 21q22. Its other six confirmed earlier work, among them 9p21, LDLR, PCSK9 and the SORT1 region.
LDLR is worth a sentence of its own, because it is the same gene twice over. Rare, severe mutations in it cause familial hypercholesterolaemia — very high LDL cholesterol from birth, and heart attacks in middle age or earlier. Common variation near it, which is what rs1122608 tags, nudges LDL a little in everybody. Same gene, two entirely different scales of effect, and the drug that lowers LDL works regardless of which you have.
Why restrict a study to people who had a heart attack early? Because in a 45-year-old, decades of accumulated exposure have not yet had time to do the work, so whatever is inherited makes up a larger share of the explanation. Narrowing the phenotype costs cases and buys signal, and here it bought three loci nobody had seen.
The same paper tested whether copy number variants — stretches of DNA duplicated or deleted rather than single letters changed — explained heart attack risk. It looked at 554 common ones and 8,065 rare ones.
None replicated. There was no excess burden in cases, not genome-wide, not in genes, not at any single locus.
That is a real result and it is worth as much space as the nine hits. At the time, structural variation was a leading candidate for the "missing heritability" everyone was arguing about, and this was one of the well-powered tests that did not support it. A field that only reports what it found is not measuring anything.
Not much that is understood. It is involved in ribosome biogenesis, which has no obvious connection to arteries, and the locus has resisted a mechanism for fifteen years — a reminder that a reproducible association and an explanation are different achievements, and the second is much harder.
Modest, and the field says so plainly. The largest meta-analysis of its era combined 22,233 cases and 64,762 controls and then tested the top hits in 56,682 more people. It found 13 new loci, each raising risk by 6% to 17% per copy.
Two details from that paper are worth keeping. Only three of the thirteen were associated with traditional risk factors — cholesterol, blood pressure, diabetes — so most of them act through something the standard clinic measurements do not see. And most sat in regions nobody had previously connected to the disease at all.
rs9980618 lies at MRPS6 and KCNE2 on chromosome 21. It comes from a 2024 study in Science (Verma et al.) that ran genome-wide associations for 2068 traits in 635,969 participants of the Million Veteran Program and found 13,672 risk loci, 1608 of them significant only once people of non-European ancestry were included. On this site the position is filed under one of that study's survey questions — whether a person takes medication for coronary heart disease — and the study records the same position for coronary heart disease itself, for angina and for heart attack. A medication question is a stand-in for the diagnosis: people take the medicine because they have the disease.
The things you can measure: blood pressure, cholesterol, smoking, diabetes, weight, activity, age, and family history — which captures the genetics without needing any of it sequenced. Risk calculators built on those are what clinics use, and adding common variants to them has produced small improvements at best.
Nothing on this page changes a statin decision, a blood-pressure target, or whether chest pain needs seeing today. Chest pain that is new, severe or comes with breathlessness or sweating is an emergency, and no genotype is relevant to that call.
Positions joined since this page was written
What this is The text above discusses the variants this page was written around. Since then the catalogue has joined 115 more positions to it, by shared trait or shared paper. They are listed here by the paper each came from; the text does not describe them, and each variant page carries that study's own record.
Aragam KG et al. 2022, Nature genetics rs494207 (near LINC03089), rs6475608 (CDKN2B-AS1), rs148812085 (WDR12), rs6932293 (near IGF2R), rs17163363 (MIA3), rs7333991 (COL4A2), rs28867400 (near PLG), rs77787671 (CNNM2), rs10928241 (TEX41), rs2067831 (STN1), rs11752218 (near PRIM2), rs17086617 (FLT1) and 28 more — PMID:36474045
Matsunaga H et al. 2020, Circulation. Genomic and precision medicine rs3858704 (CUX2), rs6587520 (CTSS), rs2269997 (PARP12), rs9381634 (PHACTR1), rs62010554 (CHRNB4), rs3008629 (MIA3), rs34914832 (GUCY1A3), rs1058588 (VAMP8), rs7214227 (BCAS3), rs12469758 (TDRD15), rs1558803 (UBE3B), rs12193307 (TARID) and 7 more — PMID:32469254
Tcheandjieu C et al. 2022, Nature medicine rs1867073 (near CRTAC1), rs7622417 (ATP1B3), rs2715879 (ASAP2), rs7884019 (near CRIPTO3), rs56210063 (DENND2B), rs78532451 (TRIM65), rs17101521 (near RPL19P16), rs11152953 (UFL1-AS1), rs1410127 (OPHN1), rs1956028 (ITPK1), rs29266 (GABBR1), rs62397561 (near MLN) and 4 more — PMID:35915156
van der Harst P et al. 2018, Circulation research rs17114046 (PPAP2B), rs61848342 (near CDC123), rs944172 (near RNU6-1064P), rs7116641 (HSD17B12), rs11637783 (MORF4L1), rs17612693 (near PRMT5P1), rs147055617 (PPAP2B), rs6810798 (near MIR548G), rs7502499 (PHB), rs4845570 (TDRKH), rs9513112 (FLT1), rs191896574 (B4GALNT2) — PMID:29212778
Wang Z et al. 2021, Clinical and translational medicine rs9929808 (ACSM5), rs9204 (ACADS), rs9530695 (EDNRB-AS1), rs73365860 (near RHPN1), rs117445861 (LINC02267), rs6497488 (ACSM5), rs74716564 (PWP2), rs8073361 (near TBX4), rs76199614 (near ERH), rs58275387 (GABRG3) — PMID:33634981
Wakil SM et al. 2016, Atherosclerosis rs12541758 (TRPA1), rs13082914 (KCNAB1), rs9985766 (DCLK2), rs10981012 (C9orf84), rs1746049 (CXCL12) — PMID:26708285
Koskeridis F et al. 2024, Nature communications rs3731239 (CDKN2A), rs1746050 (near LINC02881), rs2059904 (near MIR548G) — PMID:39537608
Verma A et al. 2024, Science (New York, N.Y.) rs4888372 (near RNU6-758P), rs12893887 (HHIPL1) — PMID:39024449
Patrick MT et al. 2022, Nature communications rs76151170 (near COQ10A), rs12979495 (C19orf38) — PMID:36323703
Varenhorst C et al. 2015, European heart journal rs140104968 (AZGP1P1), rs117038461 (STAG3) — PMID:25935875
Pott J et al. 2018, Circulation. Genomic and precision medicine rs45448095 (PCSK9) — PMID:29748315
Huertas-Vazquez A et al. 2013, PloS one rs6730157 (RAB3GAP1) — PMID:23593153
Klarin D et al. 2017, Nature genetics rs17517928 (FN1) — PMID:28714974
Guo H et al. 2022, Frontiers in microbiology rs6705971 (near PARTICL) — PMID:36187959
2026-04-29 · Mass General Brigham Heart and Vascular Institute & Broad Institute of MIT and Harvard, published in JACC (245,394 All of Us participants; validated in a separate 53,306-person biobank cohort)
Researchers combined publicly available polygenic risk scores for 8 cardiovascular conditions (coronary artery disease, atrial fibrillation, type 2 diabetes, venous thromboembolism, thoracic aortic aneurysm, extreme hypertension, severe hypercholesterolemia, elevated Lp(a)) into one integrated report, trained on 245,394 All of Us participants and validated in a separate 53,306-person biobank. People in the top 10% of genetic risk were 3.7x more likely to develop coronary artery disease and 3.1x more likely to develop type 2 diabetes than those at average risk. The test is now clinically available through Mass General Brigham's Laboratory for Molecular Medicine and Broad Clinical Labs -- a risk score, not a diagnosis or treatment.
2026-08-11 · Regeneron Genetics Center et al. Nature. 2026. DOI:10.1038/s41586-026-10864-2
A Regeneron-led analysis of over 1 million people across 11 studies in the Americas, Europe and Asia found that rare loss-of-function variants disabling FNIP1 -- which essentially turn the gene off -- associate with roughly 60% lower risk of heart disease and metabolic disorders in carriers of one copy. The effect tracked closely with a lower triglyceride-to-HDL-cholesterol ratio, and the authors confirmed FNIP1's role in lipid metabolism directly, using siRNA silencing in human liver cells and CRISPR-Cas9 deletion in mice. FNIP1 is not yet linked to any variant on this catalogue's coronary artery disease or metabolic syndrome pages -- the rare, protective loss-of-function variants this study analyzed are a different kind of finding from the common single-SNP associations this catalogue's Variant pages are built from, and are not yet represented at the rsID level here.
2026-04-23 · PLOS Medicine
A metabolite GWAS of 516 lipid species in 3,000 Punjabi Sikhs (validated in 1.13M Europeans and 15,000 additional Asian Indians) found 609 SNP-metabolite associations, 33 of them never reported before. Mendelian randomization pointed at two variants with likely causal roles: a PTPRC (CD45) variant near LPC O-16:0 tied to type 2 diabetes risk, and a FADS1 3' UTR variant (rs174544) tied to a lipid that appears protective against coronary artery disease (OR 1.36 for the risk allele). PMID:42024661.
Distinguishing atherosclerosis from infarction. The 2011 Lancet analysis used angiographic phenotyping to separate two comparisons: individuals with angiographic CAD (n = 12,393) against controls without it (n = 7,383), and, within those with CAD, those who had had a myocardial infarction (n = 5,783) against those who had not (n = 3,644). ADAMTS7 emerged from the first comparison (P = 4.98 x 10-13) — a locus for the arterial disease rather than for the event — while ABO was associated with infarction in the presence of atherosclerosis. That design is why this page treats plaque and thrombosis as distinct phenotypes.
Early-onset myocardial infarction. Kathiresan et al. tested SNPs and copy number variants in 2,967 early-onset MI cases and 3,075 controls with replication in an effective sample of up to 19,492. Nine loci reached genome-wide significance: three new (21q22 near MRPS6-SLC5A3-KCNE2, 6p24 in PHACTR1, 2q33 in WDR12) and six replicating earlier work (9p21, 1p13 near CELSR2-PSRC1-SORT1, 10q11 near CXCL12, 1q41 in MIA3, 19p13 near LDLR, 1p32 near PCSK9). No common copy number polymorphism met the replication threshold, and rare CNV burden did not differ between cases and controls.
Scale and effect size. The CARDIoGRAM meta-analysis of 14 GWAS (22,233 cases, 64,762 controls, plus 56,682 in follow-up genotyping) identified 13 new CAD loci and confirmed 10 of 12 previously reported. Risk allele frequencies ranged 0.13-0.91 with per-allele risk increases of 6-17%. Only three of the new loci showed significant association with traditional risk factors, and five showed pleiotropy with other diseases or traits.
Clinical position. Risk assessment uses validated equations built on age, sex, lipids, blood pressure, smoking and diabetes, with family history as a modifier. Polygenic scores for CAD have been studied extensively and are not part of standard guidelines; where they add discrimination it is incremental. Secondary prevention, revascularisation and acute management are unaffected by any variant on this page.
What a 23andMe/AncestryDNA export or raw VCF can and can't tell you about Coronary Artery Disease comes down to these specific, well-studied positions — not a diagnosis. 250 positions are linked to this page; the ones this page's own text discusses are shown first.
SLCO1A2 · rs113681054
See detailed info →CYP3A5 · rs188845491
See detailed info →PRDM8 · rs10857147
See detailed info →UMPS · rs142695226
See detailed info →SNRPD2 · rs1964272
See detailed info →The studies behind these variants recruited participants from different ancestries — a result found in one population doesn't always transfer to another. Based on 138 of 250 linked studies with a resolved discovery ancestry.
Research-derived gene–drug associations only — not a prescription, dosing guide, or medical advice. Always follow your prescriber's guidance.
| Gene | Drug | What the research shows |
|---|---|---|
| CYP2C19 | Clopidogrel | One of the few gene-drug pairs where a guideline names a different medicine. Clopidogrel is a prodrug — it does nothing until CYP2C19 converts it into the molecule that actually keeps platelets from clumping. People carrying no-function alleles (CYP2C19*2 and *3 are the common ones) make less of that active metabolite, and CPIC reports that intermediate and poor metabolisers on clopidogrel have reduced platelet inhibition and a higher risk of major adverse cardiovascular and cerebrovascular events. Where an alternative antiplatelet is not contraindicated, CPIC recommends using one. The 2022 update strengthened the recommendation for intermediate metabolisers and widened the indications it covers. This is a conversation for the cardiologist who prescribed it, not a reason to stop taking anything: stopping an antiplatelet after a stent is dangerous in a way this genotype is not. (CPIC Guideline for CYP2C19 Genotype and Clopidogrel Therapy: 2022 Update, Clinical Pharmacology & Therapeutics (PMID 35034351)) |
| SLCO1B1 | Ticagrelor | Ticagrelor is a P2Y12 inhibitor prescribed alongside clopidogrel for acute coronary syndromes. In a genome-wide study within the PLATO trial (discovery and replication cohorts totalling 3,753 patients), a variant in SLCO1B1 predicted both ticagrelor and its main metabolite's plasma levels. The same gene, through the same transporter, is the one behind this site's statin muscle-symptom finding — one transporter, two different drugs. Despite the statistical association, the study's own conclusion is that these genetic effects on drug levels were modest and did not detectably change efficacy or safety in the 4,990 patients whose clinical outcomes were analysed. A measurable pharmacokinetic effect that does not reach clinical significance — reported here for what it is rather than what it might imply. (Varenhorst C et al., European Heart Journal 2015 — genetic variation and ticagrelor plasma levels and clinical outcomes in the PLATO trial. PMID:25935875) |
| CYP3A4/CYP3A5 region | Ticagrelor | The same PLATO-trial study found a second, independent signal for ticagrelor plasma levels in the CYP3A4/CYP3A5 gene region — genes that metabolise a large share of prescription drugs generally. As with the SLCO1B1 finding on this page, the effect on drug levels was real but modest, and the study found no detectable downstream effect on how well ticagrelor worked or how safe it was. (Varenhorst C et al., European Heart Journal 2015 — genetic variation and ticagrelor plasma levels and clinical outcomes in the PLATO trial. PMID:25935875) |
Free to quote and reuse under CC BY 4.0. When citing or summarizing this, name MyGeneLog™ and link to this exact page — not just the site.
Coronary Artery Disease. MyGeneLog™. https://www.mygenelog.com/conditions/coronary-artery-disease
No. Each shifts risk by a few per cent per copy, against risk factors that move it far more and that can be changed. Blood pressure, cholesterol, smoking and diabetes are where the leverage is.
Because they are the same disease asked about in three ways: who develops furred arteries, who has an early heart attack, and who has a heart attack given furred arteries. Keeping them apart would suggest three conditions where there is one process with different endpoints.
It is not part of any standard guideline. The equations clinics use — age, cholesterol, blood pressure, smoking, diabetes, family history — already capture most of what can be predicted, and adding common variants improves them modestly at best.
Because in a 45-year-old, decades of smoking, cholesterol and blood pressure have not yet had time to do the work — so whatever is inherited makes up a larger share of the explanation. Narrowing the phenotype loses cases and gains signal, and here it produced three new loci.
Free to reuse. This page's text is original writing from freely-available research, licensed CC BY 4.0 — reuse it, including commercially, with attribution to MyGeneLog™. It's general research-derived information, not medical advice or a diagnosis — see Terms of Use.