Scientists use DNA to store MP3, holds 2.2 petabytes per gram
Further exploring the role biological processes could one day play in the evolution of technology, researchers from the European Bioinformatics Institute claim to have successfully encoded 154 Shakespeare sonnets and an MP3 of Martin Luther King's famed "I Have a Dream" speech into a single DNA strand -- or in other words about 739KB of information. Perhaps the most amazing part though, is researchers were able to read those files again with 100 percent accuracy. Will we eventually store our digital lives as strands of DNA?
The fact DNA can store information is of little surprise; DNA strands are essentially chemical-based instruction manuals for developing highly complex organisms with a seemingly infinite variety of permutations.
"We realized that DNA itself is a really efficient way of storing information," one researcher noted. It was this thinking led the team of scientists to consider using DNA as a digital storage medium. "So over a second beer, we started to write on napkins and sketch out some details of how that might be made to work," he continued.
In fact, DNA is so fantastically efficient, researchers believe they can cram about 2.2 petabytes of information into a single gram of DNA. "We recovered 757,051 bytes of information from 337 pg of DNA (above), giving an information storage density of ~2.2 PB/g (= 757,051/337 × 10-12)", the paper claims. That's remarkably better than today's storage technologies. By comparison, it would take several hundred 3.5-inch 4TB hard drives to match that kind of storage density.
Importantly, researchers also believe their DNA encoding scheme produces data that is reliable and long-lived. Their methods of DNA-based storage include error-correction and redundancy which protect against data loss. The big drawback though, for now, appears to be price.
According to those involved with the experiment -- using current methods for DNA manipulation -- the approximate cost per gigabyte is somewhere around $12,400/MB. "It's an unthinkably large amount of money," one researcher noted. However, his team feels that in about 10 years, DNA storage could become more cost-effective than traditional storage methods for large enterprises.
Aside from obvious cost concerns, the researchers' omission of write and retrieval speeds could prove to be another sticking point. One of the documents notes, "The experiment was not optimised for speed" but the timeline therein indicates processing times measured in "days" rather than seconds. It's unclear how much improvement can be made in this area.