All studies
DNA·November 20, 2026·6 min read

Can My 23andMe or AncestryDNA Raw Data Show My Karyotype or an Extra Chromosome Like Klinefelter, Turner or Down Syndrome

A 23andMe or AncestryDNA raw file holds genotype letters, not chromosome counts, so it cannot show or rule out Klinefelter, Turner or Down syndrome.

No. A 23andMe or AncestryDNA raw data file lists a pair of letters (a genotype) for each position it tested and nothing about how much DNA was behind them, so it cannot show your karyotype and cannot rule out an extra or missing chromosome such as Klinefelter (47,XXY), Turner (45,X) or Down syndrome (trisomy 21). A few patterns on the sex chromosomes can look like a hint, but they are far more often file quirks than biology.

What the chip measured and what the file kept

Both companies use a genotyping array: a chip with several hundred thousand probes, each reading one known position. The scanner records how brightly each probe lights up, and that brightness is how a lab tells how many copies of a stretch of DNA are present. Clinical labs use exactly that signal to find extra or missing chromosomes. The download you get keeps only the final call per position, such as AG, CC or a no-call. The brightness is gone.

Trisomy 21 shows why this matters. A person with three copies of chromosome 21 carries three alleles at a position, say A, A and G. The file has room for two letters, so the call may come out as AG, AA or a no-call, depending on how the caller handles the shifted signal. Nothing in the file says a third copy existed.

If you have a whole-genome file instead, sequencing read depth can show copy number in principle, but a downloaded VCF of variant calls is not a karyotype or a clinical test.

Three copies on the chip, two letters in the file
On the chip: three alleles
A, A and G at one position on chromosome 21, with a signal that reflects three copies
Signal intensity: reflects three copies
The brightness is what reveals extra or missing chromosomes
In the file: AG
Or AA, or a no-call, depending on how the caller handles the shifted signal
File: no intensity, no copy count
The signal is not saved, so nothing records that there were three copies

The only hints are on X and Y, and they are weak

For chromosome 21 there is no clue a reader could interpret. A third copy can shift genotype clusters and raise the no-call rate on chr21, but a slightly higher no-call rate is indistinguishable from ordinary chip noise. For the sex chromosomes there are a few clues, because X and Y follow simple counting rules that you can eyeball in the file:

  • Many heterozygous X calls in a man. A man normally has one X, so outside the tips his X should read as a single allele. 23andMe usually writes that as one letter, AncestryDNA as a doubled letter. Many heterozygous X calls would be unusual in a man, but the genotype caller may force male X calls to a single allele (or a no-call) based on the sex on the account, so an XXY man's file can look perfectly ordinary. Absence of heterozygous calls is therefore not reassurance. Check how your own file writes male X calls before reading anything into it.
  • Almost no heterozygous X calls in a woman. A typical woman is heterozygous at many X positions, so a nearly homozygous X is unusual and worth raising with a doctor. It is not a Turner test, though. Only about half of Turner syndrome is pure 45,X. Mosaic and structural forms (for example isochromosome Xq) often leave normal-looking X heterozygosity, so a heterozygous X does not rule Turner out.
  • Y calls that do not match the sex you expect. A man whose Y positions are mostly no-calls, or a woman with a scatter of Y calls. Some Y probes cross-react with X sequence, so a few Y calls in a woman's file are routine.

How many is "many"? A scattered handful of heterozygous X calls in a man is normal noise, while a consistent pattern across hundreds of non-PAR positions is not. No validated threshold exists for consumer files, so even that pattern is only a reason to ask a doctor, not a result.

Before counting, exclude the pseudoautosomal region (PAR), where a man legitimately has two copies. AncestryDNA codes X as chromosome 23, Y as 24 and the PAR as 25, so you can drop the chromosome 25 rows. 23andMe labels the same calls simply X or Y, with no separate PAR label, so exclude them by position instead: on the GRCh37 build used by most 23andMe files, PAR1 is roughly positions 60,001 to 2,699,520 and PAR2 roughly 154,931,044 to 155,260,560 on X.

Why those clues are not a diagnosis

Every one of them has a dull explanation that is more likely. No-calls are written as -- in 23andMe files and 0 0 in AncestryDNA files, and a run of them looks like missing data on any chromosome. Heterozygous-looking male X calls can also come from PAR positions you did not exclude. Probes on X and Y simply perform worse than on the other chromosomes. And in mosaicism only some cells carry the extra or missing chromosome, so any signal is diluted.

The reverse also holds. A tidy-looking X and Y proves nothing, since the caller may have forced male X calls to a single allele, and the chip intensity that would show a mosaic or a partial change was never saved. Klinefelter is common enough, roughly 1 in 500 to 1 in 1,000 males, and Turner roughly 1 in 2,000 to 2,500 females, that many people carry one without ever having been tested.

A raw data file records the letters, not the count. Only a test that counts chromosomes can answer a karyotype question.

What to ask for instead

There are two tests, and they answer different questions:

  • Karyotype. Cells are grown, the chromosomes are photographed and counted. It shows extra or missing chromosomes and large structural changes, including balanced translocations, but only down to several megabases, and it looks at a limited number of cells, so low-level mosaicism can slip through.
  • Chromosomal microarray. Reads copy number across the genome and finds gains and losses far smaller than a karyotype can see. It misses balanced rearrangements, where DNA is swapped without being lost or gained, and it can also miss low-level mosaicism.
What each test can and cannot show
Counts whole chromosomesFinds small gains or lossesFinds balanced rearrangements
23andMe or AncestryDNA raw fileSex chromosomes: weak hints onlyNoNo
KaryotypeYesOnly large onesYes
Chromosomal microarrayYes, with limits (see caption)YesNo

Scroll the table sideways for the rest →

General capabilities. Exact resolution varies by lab. For the raw file, chromosome 21 and the other autosomes show nothing; only X and Y give weak hints. Microarray caveat on counting whole chromosomes: array-CGH does not detect triploidy (SNP arrays can), and low-level mosaicism can be missed.

Both are ordered by a clinician, usually because of a reason such as delayed puberty, fertility trouble, growth or learning questions. Ask your GP, doctor or a clinical geneticist. Tell them what prompted the question and that you have a consumer DNA file.

The follow-up: what about a pregnancy or a baby

A parent's raw file says nothing about a baby's chromosomes. Most extra chromosomes, including most trisomy 21, arise by chance when an egg or sperm forms, and the chance rises with maternal age, so no variant in a consumer file predicts them. The exception is translocation Down syndrome, about 3 to 4% of cases, where a parent may carry a balanced Robertsonian translocation. A consumer file cannot see it, but a parental karyotype can, and that is the one case where a parent's own test matters. In pregnancy, blood-based screening (often called NIPT) estimates the likelihood, and a positive result is confirmed with a diagnostic test such as CVS or amniocentesis. Screening is not diagnosis, and the file you downloaded from a kit is neither.

Where a raw file is useful

The free reader at healthosx.com/dna
healthosx.com/dna
The download steps for 23andMe, on the page itself.

The download steps for 23andMe, on the page itself.

healthosx.com/dna
Ancestry across 15 regions and 157 reference populations.

Ancestry across 15 regions and 157 reference populations.

The same file does answer other questions well: CYP1A2 and your coffee response, MCM6/LCT and lactose tolerance, CYP2C19 and SLCO1B1 and how you handle certain medications, and carrier-relevant variants at named positions. HealthOSX's free DNA tool reads a 23andMe, AncestryDNA or MyHeritage file in your browser and reports traits, medication response and carrier-relevant findings, plus maternal and paternal lines. It reports no chromosome counts, sex-chromosome checks or karyotype, for the reason above, so treat it as a way to see what your file can say rather than what it cannot.

New to raw DNA files? Start with the full guide: 23andMe Raw Data Interpretation Free in Your Browser

Written by

HealthOSX Research

Keep reading