What’s inside your 23andMe or AncestryDNA raw data file

What does a raw DNA data file look like?

It is a plain-text table, usually a .txt or .csv inside the .zip you download. After a few header lines, each row is one position in your DNA, with an identifier, its chromosome and position, and your genotype:

# rsid       chromosome  position    genotype
rs4988235    2           136608646   AA
rs1815739    11          66328095    CT
rs53576      3           8804371     GG

What is an rsID?

An rsID (for example rs4988235) is a stable reference number for a specific spot in the genome, assigned by the public dbSNP database. Because it is standard, the same rsID means the same position in your file, on SNPedia, and in a research paper, which is what makes interpretation possible.

What do the two letters in my genotype mean?

DNA is built from four bases: A, T, C, and G. You carry two copies of most of your genome, one from each parent, so each position has two letters. A genotype of AA means both copies read A; CT means one C and one T. For a given SNP, which pair you carry is what determines your result for the trait linked to it.

What genome build is my raw data?

Consumer files almost always use genome build 37 (also written GRCh37 or hg19) for the position numbers. This matters only if you compare positions across tools, since build 38 uses different coordinates. The rsID stays the same across builds. We note the builds and references we use on our methodology page.

How many positions does the file include?

Roughly 600,000 to 700,000, depending on the provider and the chip version used when you tested. That sounds like a lot, but it is a small, curated slice of the roughly 3 billion bases in your genome: the chip targets positions already known to vary between people and to be informative.

Is this the same as whole genome sequencing?

No. A consumer test uses a genotyping chip that checks those few hundred thousand known positions. Whole genome sequencing reads essentially every base. So your raw data is accurate for the positions it covers but silent on everything it does not check, including most rare variants. That is the single most important thing to understand before reading too much into it.

Once the file makes sense, see how to read and interpret a single line, or browse every marker we read in the traits catalog.

More guides

• How to read and interpret your raw DNA data
• How to analyze your raw DNA data (free and paid options)
• What your raw DNA data can and cannot tell you