DNA is a written instruction, not a verdict
You asked what DNA is because a test, a relative, or a headline treated it as a verdict. It is not a verdict. It is a molecule that a cell can copy, repair, and read.
Think of a recipe, then drop the metaphor before it lies. A recipe does not bake the cake. DNA does not wander around being your personality. It is a long polymer. The useful fact is that one strand can template the other.
The real term is deoxyribonucleic acid. In almost every cell of a body, the same nuclear text is stored. What changes from tissue to tissue is which stretches are read, not which book is on the shelf.
Four letters and the pairing rule that never swaps
The text uses four bases: adenine (A), thymine (T), cytosine (C), and guanine (G). That is the entire alphabet. Meaning lives in order, the way meaning lives in the order of letters in a sentence, not in having an alphabet at all.
A pairs with T. C pairs with G. That pairing is chemistry, not a style choice: two hydrogen bonds for A-T, three for C-G, and a shape that only fits that way inside the helix. If you know one strand, you can reconstruct the other.
Each base hangs off a sugar and a phosphate. The three together are a nucleotide. Sugars and phosphates make the rails. Base pairs make the rungs.
- A (adenine) pairs with T (thymine)
- C (cytosine) pairs with G (guanine)
- A nucleotide is a base plus a sugar plus a phosphate
- The order of bases is the information
Two strands, one helix, a built-in backup
Two strands wind around each other as a double helix. They run in opposite directions. The pairing rule makes the molecule redundant: damage on one side can often be repaired from the other.
When a cell divides, the helix unzips and each strand is copied. The copies are not magic. Polymerases walk a template, make mistakes at a low rate, and correct many of them. A body is a copy of a copy of a copy, with edits.
That is why DNA can be stable for centuries in a bone and still differ between siblings. Stability and change use the same chemistry.
Nuclear DNA, mitochondrial DNA, and the cells that keep both
Most DNA sits in the nucleus, packed as chromosomes. A much smaller circle of DNA lives in mitochondria, the energy organelles. Both count. A story of DNA that forgets mitochondria is incomplete.
Nearly every cell keeps the same nuclear sequence. Mature red blood cells throw their nuclei away. Sperm and egg carry half. Tumors can rearrange the text. The textbook sentence is still useful: most cells, same nuclear DNA.
Mitochondrial DNA is present in many copies per cell. You inherit it, with rare exceptions, along the maternal line. It is DNA. It is not a nuclear chromosome.
Three billion letters, twice, and still mostly not genes
One haploid copy of the human nuclear genome is about 3.2 billion bases. Typical cells are diploid: two copies of the autosomes, so roughly twice that, plus sex chromosomes. People who say three billion usually mean one haploid set.
More than 99 percent of those bases are the same between two people. The differences are real, and they are a small fraction of the library. Most of the library is not protein-coding sequence.
A gene is a stretch the cell uses. Most DNA is noncoding: regulatory switches, repeats, RNA genes, and sequence whose jobs are still being argued about. Calling all of it junk was a shrug, not a measurement.
Copying is not destiny
Replication is how a fertilized egg becomes trillions of cells. Transcription copies a stretch into RNA. For protein-coding genes, translation reads that RNA as a protein. That flow is true and incomplete.
Cells also methylate DNA, wrap it, loop it, and ignore most of it on any given day. Those marks are annotations on the text, not a replacement for it. They can change in a lifetime. They are also not a verdict.
Identical twins share a starting text and do not live identical lives. DNA is necessary. It is not a screenplay. If you take one fact from this page, take this one: sequence is information a cell can use, not a sentence a person must serve.
Photograph 51 and the 1953 structure
In 1952, Rosalind Franklin and Raymond Gosling produced Photograph 51, an X-ray diffraction image of DNA. The X of spots was the helix made visible. Credit Franklin for that photograph.
The 1953 double-helix structure used that pattern, Chargaff's pairing rules, and model building. It explained how a chemical could carry information and copy it. History earns its keep here because the structure is the pairing rule in three dimensions.
The Human Genome Project, decades later, turned that chemistry into a reference text for a species. A structure is not a genome. A genome is not a person.
What DNA is not
DNA is not a diagnosis. A variant in a file is not a disease. A consumer chip is not a medical exam. This page is informational.
DNA is not your ancestry as a colorful pie. It is not a personality type. It is not a moral score. Companies that sell those stories are selling stories.
DNA is not the same word as gene, genome, or chromosome. Those have their own jobs. If you came here from a test result, read the molecule first, then the stretch, then the whole library.
- Not a diagnosis or a medical device
- Not a gene: a gene is a used stretch of DNA
- Not a genome: a genome is the whole library
- Not a chromosome: a chromosome is DNA packed for moving
A public DNA fact does not require a genome upload
If you already have a gene symbol or an rsID from a paper or a file, you can look up the public record without uploading a genome. genome.sh answers that kind of question. Your sequence can stay on disk.
The lookup is still not you. It is what databases currently say about a name. Names get merged. Interpretations move. Treat the result as a public annotation, then read the next guide if you want the stretch a cell actually uses.
MedlinePlus Genetics and the CDC genetics basics pages cover the same molecule in slower language, without a sales funnel. They will also tell you this is not medical advice. So will this site.
Questions
What is DNA made of?
Nucleotides. Each nucleotide is a base (A, T, C, or G), a sugar, and a phosphate. The sugars and phosphates form the backbone. The bases pair across two strands: A with T, C with G.
What does DNA stand for?
Deoxyribonucleic acid. Deoxyribose is the sugar. The bases carry the information. Acid refers to the phosphate groups.
How much DNA do humans have?
About 3.2 billion bases in one haploid copy of the nuclear genome. Typical cells carry two copies of the autosomes, so roughly twice that, plus sex chromosomes, plus a small mitochondrial genome.
Is DNA the same as a gene?
No. DNA is the molecule. A gene is a stretch of that molecule a cell uses, often to make RNA and sometimes a protein. Most DNA is not a gene.
Where is DNA in the body?
Most of it is in the nucleus of cells, packed as chromosomes. A small circle of DNA is in mitochondria. Mature red blood cells have no nucleus and therefore no nuclear DNA.
What do A, T, C, and G mean in DNA?
Adenine, thymine, cytosine, and guanine. They are the four bases. A pairs with T. C pairs with G. The sequence of those letters is the information.
Who discovered the structure of DNA?
The 1953 double-helix model explained the pairing rule in three dimensions. Rosalind Franklin's Photograph 51 made the helix visible as an X of spots on film. Credit that photograph.
Can DNA change during your life?
Yes. Cells copy DNA with rare mistakes. Sunlight, some chemicals, and cell division can add variants in particular tissues. Those somatic changes are not usually passed to children. Inherited DNA is the text in egg and sperm.