Click image to open full size
Biology
Ancestry DNA Testing
How 23andMe and AncestryDNA Work
Related Tools
Related Labs
Related Worksheets
Ancestry DNA testing uses small differences in your DNA to estimate where parts of your family history may connect to people living in different regions today. Companies such as 23andMe and AncestryDNA do not read your whole genome for this service. Instead, they scan hundreds of thousands of selected genetic markers and compare your pattern to large databases. This can reveal relatives, migration clues, and inherited segments shared with other testers.
Key Facts
- Most consumer ancestry tests use SNP arrays that read about 600,000 to 700,000 selected genetic variants.
- A SNP is a single DNA letter difference at one position, such as A instead of G.
- Humans share about 99.9% of their DNA sequence, so ancestry tests focus on the small fraction that varies.
- Shared DNA percentage is estimated from matching DNA segments, and closer relatives usually share more DNA.
- Ethnicity estimates compare your SNP pattern to reference populations, so results can change when databases and algorithms improve.
- A child inherits about 50% of autosomal DNA from each parent, but not exactly 25% from each grandparent because recombination is random.
Vocabulary
- SNP
- A SNP, or single nucleotide polymorphism, is a DNA position where people commonly differ by one genetic letter.
- SNP array
- A SNP array is a lab chip that tests many chosen SNP positions across the genome at once.
- Reference population
- A reference population is a group of people with known family origins used as a comparison set for ancestry estimates.
- Autosomal DNA
- Autosomal DNA is DNA found on the 22 pairs of non-sex chromosomes and is inherited from both parents.
- Recombination
- Recombination is the process that shuffles chromosome segments when eggs and sperm are made.
Common Mistakes to Avoid
- Treating ethnicity percentages as exact facts is wrong because they are estimates based on reference datasets and statistical models.
- Assuming a test reads every DNA letter is wrong because most ancestry tests use SNP arrays that sample selected positions rather than sequencing the whole genome.
- Thinking a region label proves nationality is wrong because genetic similarity reflects shared ancestry patterns, not modern borders or legal identity.
- Ignoring privacy settings is risky because DNA data can reveal information about relatives and may be shared, uploaded, or accessed in ways the tester did not expect.
Practice Questions
- 1 A SNP array reads 650,000 variants. If a whole genome has about 3,000,000,000 DNA base pairs, what percent of positions are directly tested?
- 2 Two people share 12.5% of their autosomal DNA. If close relationship categories include parent-child at about 50%, grandparent-grandchild at about 25%, and first cousins at about 12.5%, which category best fits?
- 3 Explain why two siblings can receive different ethnicity estimates from the same testing company even though they have the same parents.