Scientists develop DNA technology in data storage breakthrough

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Scientists have developed a new technology for using DNA genetic code to store data, boosting the quest for powerful solutions to save ever growing volumes of digital information more cheaply and sustainably.

DNA is seen as a potential salvation for data-dependent economies because it is stable and just one gramme of it can theoretically store the equivalent of about 10mn hours of high-definition video.

Now researchers from US, Chinese and German institutions have used a simple chemical reaction to mimic the binary system of traditional computers, enabling imprinting of the information on to DNA with good accuracy and much faster than traditional methods.

The promising new technology, reported in a Nature paper on Wednesday, should enable “real-world applications” of DNA storage that place less strain on electricity and other resources, said Long Qian, a co-author of the work.  

“Current data storage technologies simply [cannot] afford to store and preserve the huge amounts of data we are collecting and producing every day,” said Qian, a researcher in the Center for Quantitative Biology at China’s Peking University. “If the data is going to be stored for more than 50 years . . . preserving data in DNA will be cheaper than using and maintaining hard drives.”

Previous attempts to save information on synthetic genetic code have been time-consuming, expensive and vulnerable to errors.

In this project, the scientists exploited a natural chemical process known as methylation to modify biological building blocks in DNA known as bases. Since this meant the bases were either methylated or unmethylated, this gave them two possible states for encoding information — like the binary values of 0 and 1 used by computers.

A potential advantage of methylation is its simplicity relative to conventional methods of DNA data storage, which involve building up ever-growing amounts of new genetic code. The researchers used their technique to store images such as a colour picture of a panda and a rubbing — a reproduction of the texture of a surface — of a tiger made during China’s Han dynasty.

If existing DNA data storage techniques were like tying knots in rope to record information, the new approach was the equivalent of moveable type printing, Qian said. The researchers estimated that their encoding strategy could potentially prove 10,000 times faster than existing methods at a tiny fraction of the cost, she added.

“The process is more like ‘a printing press’ and less like ‘leaving a trail of breadcrumbs’. The hope is to make things small, fast, durable, non-damaging to the environment and inexpensive,” said Nick Goldman, a senior scientist at the European Bioinformatics Institute, adding more evaluation of the system’s speed and costs would be required.

Fuelled by increasing use of artificial intelligence, the amount of data generated annually now runs well into zettabytes, or trillions of gigabytes. This is increasing pressure on digital storage capacity, as well as stoking enormous demand for electricity from the world’s data centres.

Some companies such as Amazon, Google and Microsoft are pursuing nuclear power supply deals to meet their cloud data needs.

The new DNA data storage technique had the “potential to bypass the time and cost limitations of conventional approaches”, according to an accompanying commentary also published in Nature.

But the method still faced questions, said commentary authors Carina Imburgia and Jeff Nivala of the University of Washington. These included the long-term stability of the chemically-altered bases and the complex processes needed to copy and — in some circumstances — read them.

“The overall cost of the new system exceeds that of conventional DNA data storage and of digital storage systems, limiting immediate practical applications,” their commentary said. “Addressing the outstanding issues will help to make [the] system more scalable and practical for a broad range of uses.”