Slashdot Mirror
UW, Microsoft Successfully Encoded 200MB of Data Onto Synthetic DNA Molecules (seattletimes.com)
An anonymous reader writes from a report via The Seattle Times: Researchers from Microsoft and the University of Washington said Thursday that they had successfully encoded about 200 megabytes of data onto synthetic DNA molecules. The information included more than 100 books, translations of the Universal Declaration of Human Rights, and a high-definition music video from the band OK Go. Previously, the record was 22 megabytes encoded and decoded on DNA, said the researchers. Microsoft's lead researcher on the project, Karin Strauss, said DNA storage of the type demonstrated in the UW lab could, theoretically, store an exabyte (one billion gigabytes) of data in about one cubic inch of DNA material. "Our goal is really to build systems to show that it is possible," she said. DNA is also very durable. If stored in the right conditions, data encoded on DNA could be readable for thousands of years, compared to typical hard disks or flash drives that can fail in a few years.
The Fifth Element
Why do people always feel the need to give such ridiculous longevity estimates? Especially when you factor in the real world, that sort of longevity simply doesn't happen unless you're etching into a plate of metal (gold?) that doesn't corrode readily.
If you put your data into an organism's "junk DNA", then the data will last forever... (barring mutations)
Sounds like copyright infringement.
Maybe the BSA should get on the case.
Now, we are going to see DNA-based computer virus!
Cost of writing information: 200 MB data = 20 Mbp DNA = $200,000 USD @ $0.01 /bp. The cost was $0.30/bp for the past decade, and it's recently dropped to $0.07/bp for short fragments with the latest scale-down improvements. Some serious scale-down innovations need to occur before they drop further -- DNA synthesis costs do NOT follow a Moore-like law because DNA oligos are still produced via chemical reactions on solid supports and scale-downs have led to poor kinetics and poor yields. Cost reductions are possible, but they need to drop by a factor of 100X to even remotely compete with silicon or magnetic tape.
Cost of reading information: 200 MB data = 20 Mbp DNA = 1 MiSeq next-gen sequencing run = 150 nt read length x 10 million reads = 75X coverage = about $1100. So your reads are not cheap, and are only performed once -- ie, DNA storage could only be used as long-term write-once, read-rare storage.
Information Lifetime: "Under the right conditions, Ceze said, data encoded on DNA could be readable for thousands of years." This is a laughable statement. Under the right conditions, *any* storage medium is readable for thousands of years. So let's be serious about this. There are many real-world conditions where DNA molecules will have a lifetime of days, or even hours.
For example, if your spit or sweat accidentally mix with the DNA, some enzymes (called DNAses) could contaminate your DNA. Guess what happens? Chop chop chop chop chop. That's your information going bye bye! It doesn't have to be human spit/sweat, *any* biological contamination (bacteria, protozoa, etc) will see your DNA storage as a tasty treat to eat. Yum yum yum yum! Bye bye information! What about DNA storage in space? Well, that will certainly need some heavy duty radiation shielding because *surprise* just about every form of high energy radiation will break or mutate DNA -- UV, X-rays, gamma, etc. So even sending DNA into space for storage doesn't guarantee unlimited longevity.
It's also quite amusing that DNA (as a polymer) is not that dissimilar to another highly stable, recalcitrant polymer called cellulose, otherwise known as PAPER. Under the right conditions, PAPER *is also* readable for thousands of years. Now, paper isn't perfect, but I would choose paper over DNA for long-term storage in every way. Vacuum seal it. Firepoof it. Done. Cost to read? $0.