Heterozygous vs Homozygous: Reading the Two Letters

What AG versus AA means in a raw DNA file, why the order is irrelevant, what single letters on X, Y and MT mean, and why dominance is not about the letters.

Open a raw DNA file and every line ends the same way: two letters. AG, CC, TT, sometimes --, and near the end a run of single letters. Those pairs are the entire content of the file - everything else is a label saying where each pair came from. Understanding what a pair means is the one skill that makes the rest of your data readable.

Two copies of everything

You inherited one full set of chromosomes from each parent, so at almost every position in the genome you carry two copies. A SNP is a position where people differ by a single letter, and your genotype at that SNP is simply the letter on each copy. Two letters, one from each parent.

  • Homozygous means the two copies match: AA, GG, TT, CC.
  • Heterozygous means they differ: AG, CT, and so on.

That is the whole vocabulary. The terms sound technical, but they only describe whether the two letters are the same.

Why the order does not matter

AG and GA are the same genotype. The file cannot tell which letter came from which parent - that would require phasing, which chips do not do - so the two letters are just a set, usually written in alphabetical order. If you compare files from two services and one says AG where the other says GA, nothing has changed. Tools that match genotypes against reference tables have to account for this, which is one reason a naive string comparison of two files gives misleading mismatch counts.

Reference, variant, major, minor

For every SNP, one letter is the reference allele - the one in the reference genome - and the other is the variant (or alternate). Separately, one is the major allele - the more common one in a population - and the other the minor. These pairs usually line up but not always: at rs4988235, the variant that lets adults digest milk is the minor allele worldwide but the major allele in northern Europe. Neither “reference” nor “major” means “normal”; the reference genome is just one arbitrary human sequence. Our piece on minor allele frequency goes into this.

What a heterozygous result means for a trait

It depends entirely on the trait, and this is where people go wrong. The letters say what you carry; they say nothing about how the two copies interact.

  • For some variants one copy is enough to have the full effect. Lactase persistence works this way: one copy of the persistence allele at rs4988235 keeps the enzyme switched on.
  • For others the effect needs two copies, and heterozygotes are essentially unaffected. Hereditary hemochromatosis behaves like this: carriers of one C282Y allele at rs1800562 almost never develop iron overload, while two copies carry real risk.
  • Many are in between, with heterozygotes showing a partial or intermediate effect - the enzyme activity numbers for MTHFR 677 are the classic example.

“Dominant” and “recessive” are descriptions of these patterns, applied trait by trait; they are not properties of the letters. A single genotype can be “dominant” for one trait and “recessive” for another if the same SNP influences both.

Single letters: X, Y and mitochondria

Near the end of most files the pairs give way to single letters. That is not an error. Males have one X and one Y chromosome, so for most positions on either they carry only one copy - the technical term is hemizygous - and the file reports one letter. Mitochondrial DNA is inherited only from the mother and comes in one version, so it is also reported as a single letter for everyone. Females have two X chromosomes and get pairs there as usual.

Some services pad these to two letters (AA for a single A), which looks like a homozygous result but means the same thing. Our tour of the 23 chromosomes explains the different regions in more detail.

Compound heterozygous

One more term you will meet in medical genetics: compound heterozygous describes someone who carries two different faulty variants in the same gene, one on each copy. Each individual SNP looks heterozygous, but together they leave no working copy. A raw file cannot confirm this, because it cannot tell whether the two variants sit on the same chromosome or opposite ones. That limitation matters for interpreting anything in the recessive category.

A five-second checklist

When you look up a SNP in your file:

  1. Find the two letters. Ignore their order.
  2. Check which letter is the variant, and on which strand the file reports it - see forward and reverse strand, because the same variant can appear as A in one file and T in another.
  3. Look up how many copies matter for the trait you care about. The Genespiral marker pages list what each genotype is associated with, so you do not have to work this out from scratch.

This article is educational only and is not medical advice.

References

Further reading