New DNA Storage System Can Store 490 Exabytes Per Gram and Allows Data to Be Rewritten
By 2020, the amount of digital data produced will total 40 trillion gigabytes! Yet that mind-blowing amount of information can be crammed into just 82 grams of DNA.
For that, we can thank Professor Olgica Milenkovic and the many other researchers who have made DNA data storage a reality. Milenkovic, who is based out of the University of Illinois, believes that DNA-based storage will be the primary storage mechanism of many archival systems in the near future.
"The media is extremely durable and has exceptionally high storage density," she says.
Exceptional is a bit of an understatement. While the hard drives of many desktop computers sold today can store one terabyte of information, three years ago, researchers at Harvard created a technique that could store 700 terabytes on a single gram of DNA. A year later, another team raised the capacity to 2200 terabytes.
Just last week, however, Milenkovic and her team detailed a new system capable of storing 490 exabytes on a single gram, which is equal to 490 billion gigabytes! Besides smashing the previous storage record, her technique also permitted data stored on DNA to be selectively accessed and rewritten, two huge advances over previous DNA storage methods.
With DNA storage, data is first translated to binary (1s and 0s), then in-turn converted to DNA bases (A, G, T, and C). Once the information is laid out, DNA is synthesized to match the data. To read the information contained inside, scientists simply sequence the DNA and convert the data back to binary.
Milenkovic and her team synthesized DNA with 1000 bytes per strand. Each strand was tagged with two address sequences at each end. These sequences allowed the team to identify specific DNA strands, enabling them to access specific information. Once they amplified the strand they wanted, they could re-write the information contained within using conventional DNA editing techniques.
The researchers tested their technique on Wikipedia.
"We encoded parts of the Wikipedia pages of six universities in the USA, and selected and edited parts of the text written in DNA corresponding to three of these schools," they wrote.
While Milenkovic's DNA storage method improves on prior techniques, it is also much more expensive. Encoding and storing 17 kilobytes (KB) of data cost $4,023. A previous technique stored 739 KB of data for $12,600.
This comparison points out a glaring problem with DNA storage.
"The costs of synthesizing DNA (i.e. recording the information) are prohibitively high at the moment to allow this technology to be competitive with flash or other memories," Milenkovic told RCS in an email.
Costs are declining extremely fast, however.
"Within the last seven months, the cost of synthesizing 1000 bps blocks reduced almost 7-fold," the researchers wrote.
When DNA data storage costs drop significantly enough, the method may be of use to large governmental, scientific, or historical organizations. The Large Hadron Collider, for example, creates 15 petabytes of data each year!
But could DNA storage ever find a home in personal computers?
"I do not see how one could directly connect classical computers with DNA storage media at the moment, as one needs to do some processing on the DNA output to make it readable by a computer," Milenkovic told RCS. "But this processing may be incorporated into new generations of DNA sequencers - I am not aware of anyone working on this subject at the moment, though."
Source: Tabatabaei Yazdi, S. M. H. et al. A Rewritable, Random-Access DNA-Based Storage System. Sci. Rep. 5, 14138; doi: 10.1038/srep14138 (2015)