# DNA data storage: Science

Source page: https://www.materialinsights.org/technologies/dna_data_storage/tech-science/

#### Principles

The science it rests on, with why each principle matters to it
 Relies on

- [Structure of DNA](https://www.materialinsights.org/science/dna_structure/) Data is written as the order of the four bases along the strand, which the double helix keeps stable and which sequencing can read ^[1].
- [Error-correcting codes](https://www.materialinsights.org/science/error_correcting_codes/) Synthesis and sequencing make errors and lose strands, so codes add redundancy; the 2012 book came back with errors in only 10 bits out of 5.27 million ^[2].
 Limited by

- [Information theory](https://www.materialinsights.org/science/shannon_information/) Information theory sets the most bits each base can carry once errors and lost strands are allowed for, the capacity worked out by Shomorony and Heckel ^[1].

as of 2026-11-11

#### Discoveries

The discoveries that made it possible: when, by whom and what was found

| Year | Who | What |
|---|---|---|
| 2012 | George Church's group, Harvard | About a megabyte, a whole book, stored in DNA and read back ^[1] |
| 2015 | Grass and colleagues | Showed that data in DNA can last for thousands of years ^[1] |
| 2017 | Yaniv Erlich and Dina Zielinski | Showed that practical DNA storage can reach very high information density ^[1] |

as of 2026-11-11

#### The science explained

The science behind it explained for a curious reader

A computer file is a long string of 0s and 1s. DNA is a long string of four letters, A, C, G and T, so a file can be spelled out in DNA, each letter standing for a short group of bits. The difficulty is that machines can only build short strands, and in a drop of liquid the strands float in no order. So the file is chopped into many short sentences, each starting with a page number, an index ^[1]. Reading is like pulling random pages out of a tub: to be sure of getting every page, many are read, and error-correcting codes fill gaps. This worked well enough in 2012 for a book of 5.27 million bits to come back with 10 bits wrong ^[2].

as of 2026-11-11

#### Sources

1. [DNA-Based Storage: Models and Fundamental Limits](https://arxiv.org/pdf/2001.06311) Ilan Shomorony and Reinhard Heckel, arXiv, 2020
2. [Reading and Writing a Book With DNA](https://spectrum.ieee.org/reading-and-writing-a-book-with-dna) IEEE Spectrum, 2012

## Other sections

- [Summary](https://www.materialinsights.org/technologies/dna_data_storage/index.md)
- [Overview](https://www.materialinsights.org/technologies/dna_data_storage/tech-overview/index.md)
- [Evolution](https://www.materialinsights.org/technologies/dna_data_storage/tech-evolution/index.md)
- [Materials](https://www.materialinsights.org/technologies/dna_data_storage/tech-materials/index.md)
- [Energy footprint](https://www.materialinsights.org/technologies/dna_data_storage/tech-energy-footprint/index.md)
- [Industries](https://www.materialinsights.org/technologies/dna_data_storage/tech-industries/index.md)
