The SNP cheat-sheet

This is the switch that decides whether you make a working FUT2 enzyme. The G allele builds a functional enzyme (secretor). The A allele inserts a premature stop, so the enzyme comes out truncated and dead (non-secretor). Two A copies means no working enzyme at all. Note on reading your own data: this site is reported on one DNA strand, but a 23andMe or AncestryDNA export may print the complementary letters for this position (C instead of G, T instead of A). The biology is identical. If your file shows C/C, T/T, or C/T, read those as G/G, A/A, and G/A here.

GG

Secretor. Both copies of FUT2 work, so you make the enzyme normally. Tends to go with a more average total B12 reading. About 80% of Europeans carry at least one G.

GA

One working copy, one broken. Still a secretor in practice, since one good copy is enough to make the enzyme. B12 reading lands in the typical range.

AA

Non-secretor. Both copies carry the stop, so you make no functional FUT2 enzyme. Studies link this to a total serum B12 reading about 16 to 18 percent higher than secretors, but the extra is mostly the parked haptocorrin-bound form, not the active form your cells use. A high number here doesn't guarantee you're well-stocked.

The variant everyone studies. The T allele builds an enzyme that's slower and falls apart in heat.

GG

Typical activity. Nothing to change here.

GA

One copy. Roughly a third less enzyme activity.

AA

Two copies. Activity drops a lot, and this is where switching to methylfolate makes the most sense.

The second common one. Gentler by itself, and mostly studied next to C677T rather than alone.

TT

Typical activity.

TG

One copy. Activity is a little lower.

GG

Two copies. Still only a little lower.

The G allele tracks with slower B12 regeneration.

AA

Normal recycling speed.

AG

One copy. A modest dip.

GG

Two copies. The slowest of the three, so your B12 status matters most here.

The T allele means you build less phosphatidylcholine on your own, so more of your choline has to come from the plate.

CC

Normal choline production.

CT

Lower production, so you need more choline from food.

TT

The lowest production, and where the choline argument is strongest.

SLC19A1 sits on the minus strand of chromosome 21, so VCF allele labels and coding-strand notation run in opposite directions. T (VCF reference) = c.80A on the coding strand; C (VCF alternate) = c.80G. In the one population study with data on rs1051266, women homozygous for c.80G (that is, CC in VCF terms) had lower RBC folate than women carrying one or two A alleles. The variant accounted for roughly 5% of the RBC folate variation among women. No association appeared in men, and serum folate and homocysteine were unaffected in either sex.

TT (homozygous reference; coding AA)

Associated with higher red blood cell folate in women in the studied population. The A allele looks like the higher-folate form, at least in one Northern Irish cohort. Not a diagnosis of anything good, just a modestly favorable signal on one biomarker.

TC (heterozygous; coding AG)

Women with one copy of each allele had higher RBC folate than CC women, similar in direction to TT. Treat this as a single data point from one study. The difference is real enough to be statistically significant, but the biology here is not fully mapped.

CC (homozygous alternate; coding GG)

Lower RBC folate in women. This is the at-risk group in the one study that measured it. Even so, the SNP explains only 5% of variance in that biomarker, so your actual folate intake is doing far more work than your genotype.

rs7501331 is the workhorse variant here, the one with the cleanest human data. It changes amino acid 379 from alanine to valine, which is why you'll see it written as A379V. The T allele is the slow one. Heads up on letters: this comes from dbSNP, and a 23andMe or AncestryDNA export reads whichever strand the chip uses, so yours may print the complementary letters instead (a C shows as G, a T shows as A). Match by pattern, not by assuming the letter is wrong.

CC

Typical converter. Both copies are the reference allele, so your BCO1 enzyme handles beta-carotene at full pace. You can generally get enough vitamin A from a normal mix of plant and animal foods.

CT

One slow copy. Conversion sits somewhere in the middle. In the human study, carrying the 379V (T) allele was tied to a lower ability to turn beta-carotene into vitamin A, so plant sources work a bit less efficiently for you.

TT

Two slow copies. This is the lowest-conversion version of this variant. The 379V allele was linked to a 32% drop in beta-carotene-to-vitamin-A conversion in female volunteers, and unconverted beta-carotene tends to run higher in your blood. Preformed retinol is the more reliable vitamin A source.

rs12934922 changes amino acid 267 from arginine to serine (R267S). On its own the evidence for a big effect is thin. Where it earns its place is in combination: people carrying the slow alleles of both this SNP and rs7501331 were the worst converters in the study, well below people carrying just one. So read this row alongside the first. Same letter caveat applies: on a 23andMe-style export an A may print as a complementary T, so match by pattern.

AA

Reference at this spot. No added slowdown from this variant. Your overall conversion still depends on what you carry at rs7501331.

AT

One copy of the 267S (T) allele. Minor effect by itself. It matters most when paired with a slow rs7501331 genotype, where the two together stack up.

TT

Two copies of the 267S allele. Alone this had little measured effect, but combined with the slow rs7501331 variant it marked the largest conversion drop in the study (about 69% lower). Check both variants together to read your real conversion picture.

The G allele tracks with lower 25(OH)D and a higher shot at deficiency. Reading note: this site is reported here on the forward strand (alleles A and G). A 23andMe or AncestryDNA export may print the complementary letters T and C instead, where C lines up with G (the lower-vitamin-D version) and T with A. Match by the variant, not the exact letter.

AA

Typical vitamin-D activation.

AG

One copy. Levels tend to run a bit lower.

GG

Two copies. Lowest of the three, and the genotype where a blood test earns its keep most.

The A allele is linked to weaker vitamin-D transport.

AA

Two copies. The lowest transport of the three, and the genotype where checking your level pays off most.

AC

One copy. Transport runs a bit lower, so deficiency is more likely.

CC

Typical transport.

A single-letter change in SLC23A1 that slightly lowers how efficiently the SVCT1 transporter moves vitamin C into your blood. The T allele is the less common, lower-transport version. Note: this site is read on different strands by different labs, so a 23andMe export may print the complementary letters G and A instead of C and T (C reads as G, T reads as A). Same variant, mirrored letters.

CC

Two standard copies. Your SVCT1 moves vitamin C the usual way, so on a given diet your plasma level lands in the typical range. The most common result.

CT

One standard copy and one lower-transport copy. On the same diet you tend to sit somewhat lower than CC. In the study below, each lower-transport copy tracked with about 6 micromol/L less plasma vitamin C, so a single copy is a modest shift. Food still matters most; this just nudges your baseline.

TT

Two lower-transport copies. The least common result, and the one that tracks with the lowest circulating vitamin C for a given intake. All the more reason to keep produce steady rather than sporadic.

The G allele tends to track with slightly lower iron and hemoglobin.

AA

Typical iron and hemoglobin tendency.

AG

One copy of the G allele, so a slightly lower lean.

GG

Two copies of G, the lowest lean of the three, though your labs still have the final say.

This one sits right where the receptor protein starts being built, so it actually changes the protein. In a year-long trial on a fixed daily dose of vitamin D, the genotype groups raised their blood levels by different amounts, and one group came out clearly at the bottom.

AA

Largest rise among the groups. Sits at the opposite end of the gradient from the lowest responders, so on the same daily amount this group tends to move its blood level the most.

AG

In between. One copy of each, an intermediate response.

GG

Smallest rise per dose. In the trial this was the lowest-responding group on the same daily amount and may need a higher dose to reach the same blood level. Educational only, not a dosing instruction.

This one sits in a non-coding stretch, so it doesn't change the receptor protein itself. It's a marker that's been linked to baseline vitamin D differences, but studies disagree on the direction from one population to the next, so we won't hand you a dosing rule from it.

CC

Common version. Linked to vitamin D variation in some studies, but the direction isn't consistent across groups.

CT

One copy of each. Intermediate, with the same caveat about mixed evidence.

TT

The other version. Some studies tie it to lower baseline vitamin D, others to higher; the signal isn't settled.

A promoter variant that alters how saturated fat intake influences APOA2 gene expression and body weight regulation.

CC

Higher body weight risk specifically when saturated fat intake is high; epigenetic changes reduce APOA2 expression.

CT

Intermediate response; effects depend on specific saturated fat levels.

TT

Lower body weight risk relative to CC carriers when saturated fat intake is high; maintains higher APOA2 expression under high SFA conditions.

The variant (G) allele is tied to a bigger triglyceride jump after meals heavy in dietary fat.

GG

Two copies of the higher-response allele. This is where the omega-3 and lower-refined-carb case is strongest.

GA

One copy. Your triglycerides tend to react more than average.

AA

A typical triglyceride response.

This is the position that creates the e4 version of APOE. A C here is the e4-defining letter; a T is the non-e4 (e2/e3) letter. Note: a 23andMe or AncestryDNA export reads the other DNA strand for this spot, so it may print these as G (for C) and A (for T). Match by position, not just the letter.

TT

No e4 from this spot. Your saturated-fat-to-LDL response is the typical, less reactive one. Pair this with rs7412 to get your full e2/e3 type.

TC

One e4 copy. You carry e4. In diet studies, one e4 copy is enough to make your LDL more responsive to saturated fat, up when it's high, down when you cut it.

CC

Two e4 copies (e4/e4). The most fat-responsive type. Your LDL tends to move the most with changes in saturated fat intake.

This is the position that creates the e2 version. A T here is the e2-defining letter; a C is the non-e2 (e3/e4) letter. Note: on a 23andMe-style export this spot may print as A (for T) and G (for C) because it reads the complementary strand. Read the two spots together to land your final type.

CC

No e2 from this spot. Combined with the other SNP, you're an e3 or e4 type, the more saturated-fat-responsive end of the range.

CT

One e2 copy. The e2 version binds the liver receptor weakly and generally tracks with lower LDL. One copy nudges your profile toward the less-reactive side.

TT

Two e2 copies (e2/e2). The least common type. Cholesterol handling works quite differently here, so general APOE rules of thumb don't map cleanly; worth a conversation with a clinician if this is you.

The T allele tracks with reduced FADS1 (delta-5 desaturase) activity, the step that builds longer-chain fats like arachidonic acid, EPA and DHA from shorter ones. The effect is clearest on arachidonic acid; the shift in EPA and DHA is real but smaller and varies between studies.

CC

Typical desaturase activity.

CT

Somewhat lower activity.

TT

Lowest of the three. If your omega-3s come mostly from plant ALA, preformed EPA and DHA is the more dependable bet.

The A allele is the inducible "fast" version. Carry a C and your clearance slows down.

CC

Slow clearer. Caffeine lingers, so keep it earlier and lighter.

CA

In between. Slower than two A's, faster than two C's.

AA

Fast clearer. Caffeine moves out quickly.

The A allele is associated with higher BMI and increased food intake, while the T allele is linked to lower intake.

AA

Higher baseline appetite and energy intake; ~0.39 kg/m² higher BMI on average.

AT

Intermediate risk profile between AA and TT.

TT

Lower baseline appetite and energy intake; reference group for lower BMI risk.

T is the version that adds risk. Each copy you carry stacks the effect a bit higher.

CC

Two C's. Nothing extra coming from this variant.

CT

One T. Your type-2 diabetes risk ticks up a little.

TT

Two T's. This is the strongest signal a common variant gives, so your blood-sugar habits carry the most weight.

The proline-to-leucine swap in glutathione peroxidase 1. In your raw export the alleles read as G and A. The research literature writes the same spot as C (proline, the common, more responsive form) and T (leucine, the variant). Your 23andMe or AncestryDNA file may print the complementary letters, so don't be thrown if you see C/T instead of G/A.

GG

Two copies of the common (proline) form. In the study below, GPx1 activity tracked selenium most tightly in this group: the enzyme ramps up well when selenium is around. Standard selenium needs apply.

GA

One common copy, one variant (leucine) copy. An in-between response: the enzyme still answers to selenium, just less briskly than GG. Most people carry at least one copy of each.

AA

Two copies of the variant (leucine) form. In the study, this group showed the weakest, non-significant link between selenium and enzyme activity, which points to a blunted response. Not a deficiency, not a diagnosis. Just a flatter return on the same selenium.

The G allele builds proline at position 187 and gives you a fully working enzyme. The A allele swaps in serine, and that version of the protein is unstable. Your cell breaks it down fast, leaving much less CoQ10-recycling activity. Two A copies leaves only a small fraction of normal enzyme activity. One note: a 23andMe export reads the other DNA strand, so it may print these as C and T instead of G and A. C lines up with G (the working version), T lines up with A (the variant).

GG

Two working copies. Your NQO1 recycles CoQ10 and vitamin E at full speed. The typical, higher-activity result.

GA

One working copy, one variant copy. Activity sits in the middle, so recycling runs lower than GG but holds up better than AA.

AA

Two variant copies. The serine version of the enzyme gets degraded quickly, leaving only a small fraction of normal activity. The lowest CoQ10-recycling capacity of the three.

This is the single most studied spot in the SOD2 gene. It changes one amino acid in the targeting label that tells the cell to ship MnSOD into the mitochondrion. The dbSNP letters here are A and G. Many consumer DNA files print the complementary letters T and C instead, because the variant can be read off either strand of the DNA. If your export shows T and C, that is the same variant read the other way around: T lines up with A, and C lines up with G. The G allele (C on the other strand) builds the version that imports cleanly and gives you more active enzyme. The A allele (T on the other strand) builds the version that partly gets stuck on the way in, leaving less working enzyme where it is needed.

GG

Both copies build the high-import version of MnSOD. The targeting label holds its shape, the enzyme reaches the mitochondria efficiently, and you make the most active MnSOD of the three genotypes. This is the alanine/alanine combination.

AG

One copy of each version. You import a working amount of MnSOD, somewhere between the other two genotypes. This is the most common result for many people and is nothing to worry about.

AA

Both copies build the lower-import version. More of the enzyme gets partly stuck on the way into the mitochondrion, so you end up with less active MnSOD than a GG carrier. Reason to be steady about feeding your antioxidant defenses, not a diagnosis of anything. This is the valine/valine combination.

This is the only ACTN3 variant that matters for the power-versus-endurance story. The letter you carry tells you whether your fast-twitch fibers contain alpha-actinin-3. Heads up on reading your raw data: dbSNP records this SNP on the minus strand, so a 23andMe export may print the complementary letters (G/A instead of C/T). If you see G/A, read G where this table says C and A where it says T.

CC

Two working copies (the R/R genotype). Your fast-twitch fibers are fully loaded with alpha-actinin-3 and lean toward fast, powerful, glycolytic contractions. This is the most power- and sprint-associated version.

CT

One working copy, one switched off (R/X). You make alpha-actinin-3, just less of it. A middle-ground profile that sits between the two extremes. This is the most common genotype in most populations.

TT

Both copies switched off by R577X (X/X), so you make no alpha-actinin-3 at all. About 16% of people worldwide. Your fast fibers shift toward more efficient aerobic metabolism and resist fatigue a bit longer during sustained effort, which tilts the profile toward endurance.

Swaps one amino acid (position 16) at the start of the beta-2 receptor, changing how the receptor behaves under repeated adrenaline. Often pitched as the fat-burning variant. It mostly isn't. Here the G allele codes glycine (Gly16) and the A allele codes arginine (Arg16).

GG

Gly16/Gly16, the more common form in many groups. The more responsive setup: under direct beta-agonist stimulation in the lab, the glycine form shows the fuller rise in free fatty acids and fat oxidation. In real training, the everyday difference is small.

GA

One copy of each, glycine and arginine. Intermediate, and where a lot of people land. No reliable real-world advantage or penalty for exercise fat loss.

AA

Arg16/Arg16. In controlled studies that infused a beta-agonist and measured fat release directly, this arginine form showed a blunted rise in free fatty acids and glycerol and a bit less fat oxidation, more so in women. It is a modest cellular difference under drug stimulation, not a measured penalty you would feel during normal training.

A single C-to-T change that swaps valine (Val) for methionine (Met) at position 66 in the BDNF protein. The C allele is the valine version (full activity-dependent release); the T allele is the methionine version (the one that blunts release). The reduction generally tracks with the number of T copies you carry.

CC

Two valine copies. This is the standard version, with the most efficient activity-dependent BDNF release. Your brain packages and dispatches BDNF normally when neurons fire.

CT

One valine, one methionine. Activity-dependent BDNF release is modestly reduced compared to CC, sitting between the two homozygous genotypes. The steady background release is unaffected; only the activity-triggered burst is dialed down.

TT

Two methionine copies. This genotype shows the largest reduction in activity-dependent release. Background (constitutive) release stays normal; it's the burst tied to neural activity that takes the biggest hit.

A G-to-A change at this spot swaps valine for methionine in the enzyme. The methionine version is less stable at body temperature, so it clears dopamine more slowly. Activity steps down across the three genotypes: GG is fastest, AA is slowest, GA sits in the middle. Note on testing: rs4680 is reported here on the strand where the variant reads as G or A. A 23andMe or AncestryDNA export may print the complementary letters instead (C for G, T for A), so an AA result there could show up as TT. Same variant, just the other strand.

GG

Two copies of the valine (Val) version. This is the fast, heat-stable enzyme, so dopamine gets cleared quickly. The 'warrior' end. Lower resting prefrontal dopamine; often described as steadier under stress, less sharp on tasks that need peak focus.

GA

One Val, one Met. Intermediate enzyme activity and intermediate dopamine clearance. The most common genotype, and biochemically the middle of the road.

AA

Two copies of the methionine (Met) version. This is the slower, less heat-stable enzyme, so dopamine lingers longer. The 'worrier' end. Higher resting prefrontal dopamine; often linked to better focus when calm but more sensitivity to stress.

The T (A) allele confers lactase persistence; it's dominant, so one copy is usually enough.

AA / TT

Lactase persistent — usually tolerate dairy.

AG / CT

One persistence copy — usually still tolerate dairy.

GG / CC

Non-persistent — little adult lactase; dairy more likely to bother you.

This single letter change swaps one amino acid in the bitter-taste receptor at position 49. In the orientation we use here, the C version builds the sensitive (proline) receptor and the G version builds the sluggish (alanine) one. It travels with two nearby variants as a package, but this site does a lot of the work on its own. One caution worth knowing: this is a notoriously strand-flippable SNP. Consumer files (like 23andMe or AncestryDNA) report it with the letters C and G, but different labs read it from opposite DNA strands, so the same physical result can be printed with the taster and non-taster letters swapped relative to the rows below. Because of that, the safest way to read your result is by the taster vs non-taster interpretation your provider gives you, or a strand-aware lookup tool, rather than hand-matching the raw C/G letters from your file.

CC

Two copies of the sensitive build. You are most likely a strong bitter taster: PTC paper tastes harsh, and raw cruciferous greens can read as genuinely bitter. In a controlled taste study, CC tongues detected the bitter compound PROP at the lowest concentrations.

CG

One of each. Intermediate sensitivity. Bitter greens register, but usually as background rather than a wall. Most people fall here.

GG

Two copies of the quiet build. You are likely a bitter non-taster: PTC paper tastes like plain paper, and the bitterness in kale or sprouts mostly slips past you. Carriers of the G version needed higher concentrations before they could detect PROP at all.

This variant sits in the CETP gene and is one of the most replicated genetic associations with HDL cholesterol in the literature. The A allele (B2) links to lower CETP activity and higher HDL; the G allele (B1) to higher activity and lower HDL. One thing worth knowing: the often-cited 14% HDL difference comes from a two-SNP diplotype combining rs708272 with a second CETP variant. The single-SNP effect at rs708272 alone is directionally consistent but smaller.

GG (B1/B1)

Both copies are the B1 allele, linked to higher CETP activity and more cholesterol transferred away from HDL particles. B1/B1 carriers tend to have the lowest HDL of the three genotypes. The direction of the effect is well established; the full magnitude shows up when rs708272 is analyzed together with a nearby CETP variant as a diplotype, not from this SNP alone.

GA (B1/B2)

One copy of each. CETP activity and HDL levels land between the two homozygous groups. The most common genotype in many populations.

AA (B2/B2)

Both copies are the B2 allele. Lower CETP activity, HDL stays in circulation longer. When rs708272 is analyzed in a two-SNP diplotype with the nearby -4502C>T variant, the B2B2/TT group showed HDL about 14% higher than the B1B1/CC group in a Mediterranean cohort of over 4,200 people. That figure belongs to the diplotype, not to rs708272 by itself.

GSTM1 varies by deletion of the entire gene, so the meaningful states are 'functional' (at least one working copy) or 'null' (both copies deleted) — not a genotype like AA/AG/GG.

Functional (present)

You carry at least one working copy and make the GSTM1 enzyme.

Null (both copies deleted)

Roughly half of people, varying by ancestry. You make no GSTM1 enzyme; related glutathione S-transferases handle much of the same job.