Over the past few years, scientists have decoded the 700,000-year-old mammoth and horse genomes using DNA fragments extracted from fossils. DNA definitely lasts much longer than organisms for which it carries genetic codes. Computer scientists and engineers have long dreamed of harnessing the diminutiveness and resilience of DNA to store digital data. They want to encode all those zeros and ones into molecules A, C, G and T, which form the spiral staircase of DNA polymer - and this decade's advances in DNA synthesis and sequencing have led to a major breakthrough. Recent experiments have shown that one day we will be able to encode all of the world's digital information in a few liters of DNA - and read it again thousands of years later.
Interest from Microsoft and other tech companies is raising tensions in this area. Last month, Microsoft Research said it would pay synthetic biology startup Twist Bioscience to create 10 million DNA strands designed by Microsoft's computer scientists to store data. Leading memory maker Micron Technology is also funding DNA storage research to determine if a nucleic acid system can push the limits of electronic memory. This influx of money and interest could gradually reduce exorbitant costs and make storing data in DNA possible within ten years, the researchers say.
Humans will generate over 16 trillion gigabytes of digital data by 2017, and most of that will need to be archived. Legal, financial and medical data, as well as, of course, multimedia files. Today, data is stored on hard drives, optical disks in energy-intensive data centers the size of a warehouse. At best, this data is stored for thirty years, at worst - several. In addition, according to Microsoft Research computer architect Karin Strauss, "We are producing a lot more data than the storage industry can do, and the projections show that the gap will widen."
Now let's add DNA to all of this. It can live for centuries if kept in a cool, dry place. In theory, it can pack billions of gigabytes of data into a sugar cube. Tape, the densest storage medium available today, can hold 10 gigabytes in the same amount of space. “DNA is an incredibly dense, durable and non-volatile storage medium,” says Olgica Milenkovic, professor of electrical and computer engineering at the University of Illinois at Urbana-Champaign.
This is because each of the four building molecules - adenine (A), cytosine ©, guanine (G) and thymine (T) - occupies a cubic nanometer in volume. Using a coding system - say, in which A represents the bits “00,” C represents “01,” and so on - scientists can take the rows of ones and zeros that make up digital data files and create a DNA strand containing a snapshot or video. Of course, the real coding technique is much more complicated than we wrote for you here. The synthesis of a designer DNA strand is the process of writing data. Scientists can then read them by sequencing the chains.
Harvard geneticist George Church founded this area of research in 2012 by encoding 70 billion copies of the book - one million gigabits - in a cubic millimeter of DNA. A year later, scientists from the European Bioinformatics Institute showed that they could read, without a single mistake, 739 kilobytes of data enclosed in DNA.
Last year, several teams of scientists demonstrated fully functioning systems. In August, scientists at ETH Zurich encapsulated synthetic DNA in glass, subjected to conditions simulating the expiration of 2,000 years, and completely recovered the encoded data. In parallel, Milenkovic and her colleagues reported that six American universities had saved Wikipedia pages in DNA and - by providing the sequences with special "addresses" - selectively read and edited portions of the written text. Random access to data is critical to avoid having to "sequence an entire book to read just one paragraph," Milenkovich says.
In April, Strauss and scientists Jord Seelig and Luis Tsese of the University of Washington reported that they were able to write three image files, each several tens of kilobytes, in 40,000 DNA strands using their own encoding scheme, and then read them individually, not making mistakes. They presented their work in April at the conference of the Association for Electronic Computing. With the 10 million strings that Microsoft buys from Twist Bioscience, scientists plan to prove that DNA data can be stored on a much larger scale. “Our goal is to demonstrate a final system in which we encode DNA files, synthesize molecules, store them for a long time, and then restore them by sequencing the DNA,” says Strauss. "We start with the beats and go back to the beats."
Promotional video:
Memory manufacturer Micron is studying DNA as a post-silicon technology. The company is funding the work of Church and University of Idaho scientists to create an error-free storage system in DNA. “The rising cost of storage will drive alternative solutions, and DNA storage is one of the most promising solutions,” said Gurtei Sandu, director of advanced technology development at Micron.
Scientists are still looking for ways to reduce the number of errors in data encoding and decoding. But most of the technology is already in place. So what's stopping us from moving from shoebox-sized data warehouses to glass capsules of DNA? Cost. “The recording process is a million times more expensive,” says Seelig.
Here's why: Making DNA involves stringing nanosized molecules one by one with high precision - it's not an easy task. And although the cost of sequencing fell due to the rapidly growing demand for this service, DNA synthesis did not have a similar driver on the market. Milenkovic paid about $ 150 to create a string of 1,000 synthesized nucleotides. Sequencing a million nucleotides costs about a cent.
Interest in data storage from Microsoft and Micron may be just the momentum needed to start cutting costs, Seelig says. Clever engineering and new technologies like microfluidics and nanopore DNA sequencing that help reduce and speed up the process will also help advance. It now takes hours to sequence several hundred base pairs - and days to synthesize them - using a bunch of equipment. I wish I could do it all in a small box, otherwise the advantage of storage density would be lost.
If all goes well, Strauss envisions companies offering archival DNA preservation services for the next decade. “You can open a browser and upload files to their site or take your bytes back like you would with the cloud,” she says. Or you could buy a DNA disc instead of a hard drive.
ILYA KHEL