Data Storage in DNA Becomes a Reality

January 26, 2013 2:33 PM

"The researchers began with the computer files from some notable cultural highlights: an audio recording of MLK Jr.’s 1963 “I Have a Dream” speech, all 154 of Shakespeare’s sonnets, and, appropriately, a copy of Watson and Crick’s original research paper describing DNA’s double helix structure. On a hard drive, these files are stored as a series of zeros and ones. The researchers worked out a system to translate the binary code into one with four characters instead: A, C, G and T. They used this genetic code to synthesize actual strands of DNA with the content embedded in its very structure. The ouput was actually pretty unimpressive: just a smidgeon of stuff barely visible at the bottom of a test tube. The wow factor arose when they reversed the process. The researchers sequenced the genome of the data-laden DNA and translated it back into zeros and ones. The result was a re-creation of the original content without a single error, according to the results published in Nature on Wednesday."
posted by SpacemanStix (37 comments total)

DNA encoded porn in 5...4...3...2...
posted by Tell Me No Lies at 2:37 PM on January 26 [3 favorites]

I could have sworn I read a Bradbury short story on this very topic - creatures that were designed to have the classics encoded in their DNA. IIRC, the story ends (spoilers?) with mutations creeping in and corrupting the information.
posted by jquinby at 2:39 PM on January 26

Backups would be much more fun to make!
posted by fullerine at 2:43 PM on January 26 [11 favorites]

Unless the Martin Luther speech was encoded as a flac, I'm not interested. Lossy transcodes are for plebes.
posted by Keith Talent at 2:45 PM on January 26 [10 favorites]

2013: consumers access fractional human knowledge on cumbersome 'smartphone' devices
2020: citizens absorb the entirety of human knowledge via personalized skin rings
posted by prinado at 2:49 PM on January 26 [2 favorites]

The last author, Ewan Birney, made a blog post about this, "Using DNA as a digital archive media." Ewan has a great research method:

Where would science be without a pub?
At the end of a particularly long day, Nick and I were having a beer and talking about the need for dense, low-cost storage. We joked that of course the densest, most stable way to store digital information was in DNA, so we should just store sequenced DNA information in … DNA. We realised that this would only work if the method could handle arbitrary digital information – but was it possible? Still in a playful frame of mind, we got another drink and started to sketch out a solution.

Ewan was also my PhD adviser, and I am sure he has bought me more beers than anyone in science. Thanks, Ewan!
posted by grouse at 2:52 PM on January 26 [12 favorites]

Backups would be much more fun to make!

Hey, baby, wanna see my ... zip ... drive? I promise, it's 8 gigs.
posted by dhartung at 2:53 PM on January 26 [1 favorite]

Double?
posted by ShutterBun at 3:00 PM on January 26

Metafilter: GGAATTGCCGTACCGTACT
posted by cacofonie at 3:01 PM on January 26 [2 favorites]

Metafilter: ATGGAAACAGCATTCATACTAACAGAACGA
posted by 445supermag at 3:10 PM on January 26

It's a double, but these are actually different people doing the same thing. One thing they did which is much better than the previous article is that they used coding theory to greatly improve accuracy. In particular, they encoded the information in a way that avoids repeating the same letter, which commonly results in reading errors from current DNA sequencing technologies. This was a very simple code, however, and can be improved.

They also mention the storage density of the technology at 2.2 petabytes per gram, which I speculated in the previous thread would be the most dense that we get for quite some time. Packing more information into the same shape would greatly complicate the readers, as well as typical techniques for dealing with DNA such as amplification. So for the moment, I see 2.2 petabytes per gram as the best that humans will do until a major technology breakthrough.

Also, if you have a maker faire coming up soon, anyone can do this on their own, just order up the sequences you want from Agilent (the HP spinoff that doesn't suck). The current process was published in 2010.
posted by Llama-Lime at 3:12 PM on January 26 [1 favorite]

For comparison on the density, the best we can do right now with magnetic media is probably LTO-6 at ~12GB/g, a mere 180,000x less dense.
posted by Llama-Lime at 3:24 PM on January 26 [4 favorites]

But I don't want my tomatoes to taste like Shakespeare's sonnets! Down with GM food! Etc.
posted by neuromodulator at 3:42 PM on January 26 [2 favorites]

And the obvious next step is to bring back extinct reptile species specifically to do this.

Because who wouldn't want their own steganosaurus?
posted by mephron at 3:44 PM on January 26 [21 favorites]

It's a double, but these are actually different people doing the same thing.

Both articles seem to eventually point back to the same study reported in Nature Magazine
posted by ShutterBun at 3:44 PM on January 26

There's a third MeFi post which is about the previous study in Science from the George Church lab.
posted by grouse at 3:47 PM on January 26

Oops, somehow I missed that the "Double" link was to this same Nature publication rather than the Church Science publication. Yes, a double, in two senses. (Though on the scientific side, I think that the Church and Birney labs can call it a tie on the timing, and the Birney lab gets points for style.)
posted by Llama-Lime at 3:54 PM on January 26

I should point out the first and last authors (Goldman, Bertone, Birney) on this are all principal investigators at the European Bioinformatics Institute and the other people who worked on this are in the Goldman lab. I would characterize this as a paper from the EBI or from Goldman, Birney, and co-workers rather than from the Birney lab.
posted by grouse at 4:06 PM on January 26 [1 favorite]

I think DNA has already been storing data for quite a while now.
posted by snofoam at 4:12 PM on January 26 [3 favorites]

A bit of a derail, but it would make a great science fiction story now for someone to figure out that human DNA has a message encoded in it. The message could contain a means for contacting some alien race that engineered us.
posted by DrumsIntheDeep at 4:26 PM on January 26 [2 favorites]

The message could contain a means for contacting some alien race that engineered us.

"The Chase", ST:TNG.
posted by stebulus at 4:36 PM on January 26 [2 favorites]

Our DNA just has some dude's collection of animes and comic books, don't freak out, it is all worthless.
posted by TwelveTwo at 4:46 PM on January 26

The message could contain a means for contacting some alien race that engineered us

ALL THESE DRINKS ARE YOURS
EXCEPT OVALTINE
posted by jquinby at 4:48 PM on January 26 [10 favorites]

Huh, in magnetic media we're only about 17.5 doublings away from the information density of DNA? I would have guessed it was more than that.
posted by jepler at 4:48 PM on January 26

Double?

Would have been funnier if you said "Clone" or "Replicate". All the bioinformaticists in the house know what I'm talking about.
posted by Blazecock Pileon at 4:56 PM on January 26 [3 favorites]

… and there are 64GB micro sdxc cards. Wikipedia gives the typical weight of a micro sd card as .25 gram for a higher 256GB/gram, making it only about 13 doublings away from the figure for DNA. Hard drives are well behind at 5GB/gram (WD RE 4TB drive at 1.66lb / 753 gram).
posted by jepler at 5:09 PM on January 26 [1 favorite]

But I don't want my tomatoes to taste like Shakespeare's sonnets!

Shall I compare thee to a summer's day?
Thou art more lovely and more temperate.
But produce shouldn't be that shade of gray,
Or disconcert with built-in sell-by dates.
posted by Mr. Bad Example at 5:29 PM on January 26 [2 favorites]

#262095 (7580/8950) ↑Funny ↓Not ⚐Flag
1. The human cell contains 75 MB of genetic information
2. A sperm 37.5 MB.
3. In a milliliter, we have 100 million sperms.
On average, one ejaculation releases 2.25 ml in 5 seconds.
Using basic math we can compute the bandwidth of the human male penis as:
(37.5MB x 100M x 2.25)/5 = (37,500,000 bytes/sperm x 100,000,000 sperm/ml x 2.25 ml) / 5 seconds = 1,687,500,000,000,000 bytes/sec = 1,687.5 TerraBytes/sec
Sweet
DoS attack!!!
a bukkake would probably be a DDoS then
11 men would give 17 petabytes/secComment: ##programming on FreeNode
posted by thewalrus at 5:46 PM on January 26 [1 favorite]

I'm clicking the not funny arrow but it's not doing anything.
posted by desjardins at 5:55 PM on January 26 [2 favorites]

But I don't want my tomatoes to taste like Shakespeare's sonnets!

Yet nature is made better by no mean
But nature makes that mean. So, over that art
Which you say adds to nature, is an art
That nature makes.
posted by washburn at 6:30 PM on January 26 [1 favorite]

an audio recording of MLK Jr.’s 1963 “I Have a Dream” speech, ... and, appropriately, a copy of Watson and Crick’s original research paper describing DNA’s double helix structure.

I expect these pirates will hear from U.S. Attorney Carmen Ortiz any time now.
posted by ChurchHatesTucker at 6:44 PM on January 26 [2 favorites]

In the future pirates will encode movies into the DNA of plants, and pirated films will distribute themselves through dandelion seeds.
posted by Pyry at 7:03 PM on January 26 [4 favorites]

jquinby may be thinking of Philip K. Dick's The Preserving Machine.
posted by a person of few words at 7:41 PM on January 26 [1 favorite]

Yes! That's the one!
posted by jquinby at 4:11 AM on January 27

So, what happens if my DNA is infringing someone's copyright?
posted by Obscure Reference at 7:21 AM on January 27

How do we know that the bits of DNA we don't understand yet aren't really messages from some past intelligent race on Earth?
posted by ZeusHumms at 11:26 AM on January 27

So, what happens if my DNA is infringing someone's copyright?

You will remain under house arrest until you agree to have a Hulu banner ad spliced into your DNA.
posted by jonp72 at 6:57 PM on January 27