5 ms·

DNA may soon be used to store computer data

51 points by jkuria 8y ago

rocky1138 8y ago
This reminds me of "The Chase" from Star Trek: The Next Generation. Its plot centred around data stored in the DNA of different species across the galaxy.
presscast 8y ago
Isn't DNA orders of magnitude slower than current storage media?
- jfarlow 8y ago
  Yes. Many orders of magnitude. However, if you're just doing cold storage, the data-density of DNA is many orders of magnitude greater. With essentially sub-nanometer 3D atomic data-density you can pack a staggering amount of data in a very very small package. And though we're not quite there yet in read/write speeds at the ends of the route, the old saying rings true about the fastest way to transfer large amounts of data between the coasts - in a station-wagon full of hard drives.
  - Real_S 8y ago
    Also, consider that read-speed is rapidly improving. So for long-term storage, reading the data at a reasonable rate in the future is almost guaranteed. Hopefully write-speed will reach a similar rate of improvement. Importantly, expected corruption of the data stored in DNA is less than what one expects using other forms of data storage. So if stability of the data is a concern, DNA has unique advantages.
    - drvdevd 8y ago
      I wonder what effect the introduction of artificial nucleotides and engineered proteins will have on “DNA” storage speeds as well...
technobabble 8y ago
Does anybody recommend specific research papers on how this is done?
- jfarlow 8y ago
  - High level article [1] - A "Roadmap" from the semiconductor/data side, looking at DNA from its first principles [2] - Review combining the tech side with the bio side [3] [1] http://www.sciencemag.org/news/2017/03/dna-could-store-all-worlds-data-one-room http://www.sciencemag.org/news/2017/03/dna-could-store-all-w... [2] https://www.src.org/program/grc/semisynbio/ https://www.src.org/program/grc/semisynbio/ [see PDF on right side of page] [3] https://www.biorxiv.org/content/biorxiv/early/2017/03/07/114553.full.pdf https://www.biorxiv.org/content/biorxiv/early/2017/03/07/114...
onychomys 8y ago
It's super slow, hard to read, easily corruptible, and expensive. Literally the only advantage is storage density. But we live in a world where giant companies build immense data centers in places nobody cares about, like Nevada. So I'm not sure if anybody is actually worried about storage density.
- i_am_nomad 8y ago
  Can we maybe cut the coastal snobbery just a bit? Plenty of people care about Nevada, specifically, Nevadans. (Not sure that’s the correct demonym.)
  - choward 8y ago
    Thanks. It is the correct demonym. Just because there is a lot of unused land doesn't mean nobody cares.
- PyroLagus 8y ago
  It maybe easily corruptible, but it's super easy to copy. Nature has already solved that part for us already. If you can just have some bacteria make a million copies of your data, it doesn't matter if even several thousands of the strands become corrupted. But of course, unless you can actually read several hundred thousand strands in a reasonable amount of time, that redundancy doesn't help too much. I wonder if we could engineer bacteria to facilitate reading and writing to DNA in a more efficient manner.
- MauranKilom 8y ago
  Let's not forget one church building archiving the internet several times over.
calgoo 8y ago
"Someone creates a virus to delete or encode all your data that's stored in DNA secure storage. This then spreads and starts to encrypt the DNA sequence in a humans body. A new step in computer malware happens where malware writers start modifying people and unless you pay the 100 bitcoin to unlock your DNA in 24 hours you are dead." Its an interesting time to live in. On the one hand i love the push for all the new tech we are creating. However, at the same time, the old problems dont go away because we create new tools.
- Real_S 8y ago
  Interesting thoughts. Encrypting dna molecules is possible, but it probably wouldn't use a virus. Also, if your cell's DNA was changed like this, it would kill you very quickly, interfering with normal cellular function.
  - jobigoud 8y ago
    A better strategy would be to target the gametes and encrypt these only. Now if you want to be unsterilized you have to pay. Maybe it could a contraceptive method.
- mr_toad 8y ago
  This makes about as much sense as a computer virus that destroyed the entire OS of a machine it infected. It would never be able to propagate.
  - staplers 8y ago
    Unless it propagates before the trigger is pulled
    - thomasfedb 8y ago
      Lots of examples of this in extant biology - kill genes and mechanisms for destroying cells and DNA that are vital to the survival of multicellular organisms such as humans.
Real_S 8y ago
Also consider that cryptography can be applied directly to DNA molecules. Cryptography for genetic material: https://www.biorxiv.org/content/early/2017/06/30/157685 https://www.biorxiv.org/content/early/2017/06/30/157685
- p1mrx 8y ago
  Why would a new storage medium require a new method of cryptography? These problems seem orthogonal.
  - emmelaich 8y ago
    All algorithms have engineering trade-offs to for practical application. e.g. available memory, cpu, time. DNA is very slow but very dense.
pronoiac 8y ago
"Massive Attack’s Mezzanine “remastered” in aerosol spray can format," as DNA > The complex encoding process is explained in detail in by Dr Robert Grass of Zurich-based company TurboBeads. “This digital bitstream of the album (0s and 1s) was first translated to 901’065 DNA sequences (A, C, T and Gs), each 105 characters long”, says Grass. “The 901’065 individual sequences were then chemically synthesised resulting in a synthetic DNA sample, which fully represents the digital bitstream of the album.” > Then, in order to “guarantee information stability”, the DNA sequences were encapsulated in “synthetic glass fossils”, which were added directly to the spray can. According to Dr Grass, each can “contains at least 0.1 micrograms of the synthetic DNA, which is equivalent to 1 million copies of the album.” Playback is left as an exercise for the listener http://www.factmag.com/2018/10/19/massive-attacks-mezzanine-remastered-spray-can/ http://www.factmag.com/2018/10/19/massive-attacks-mezzanine-...
deepnotderp 8y ago
Last I looked into this (I think micron was looking into it also?) these were the major problems: 1. Orders of magnitude slower than normal memory. 2. Could be easily damaged relative to current technology. 3. Requires orders of magnitude more energy than standard storage media. 3. Is currently far more expensive than standard storage media as well. And unfortunately, due to the fundamental nature of the storage media, I don't know if the orders of magnitudes jumps that we're used to with conventional storage/ memory technology will be possible here. So the only real advantage is density, but if it's going to be used as cold storage anyways, does anyone really care compared to tape? And is this ultimately more promising than atoms on a surface (e.g. like IBM millipede)?
- pcrh 8y ago
  It also has the potential to be introduced into a living organism. In this manner it could well last for millennia by being transmitted to the organism's descendants. Assuming there is no disadvantage of carrying this DNA it could perdure over geological periods of time.
  - colejohnson66 8y ago
    Until mutations happen. But redundancy should help with that
- wuschel 8y ago
  I had similar results when looking into this - it was not competitive in cost or performance (speed, errors) a year ago.
Apocryphon 8y ago
If we're going to talk about alternative computing media, how's optical circuitry doing these days?
ImaCake 8y ago
What often seems to be overlooked when talking about using DNA for storage is that microorganisms want to eat it for their own uses. DNA in the environment isn't just ignored, it is quickly degraded by excreted enzymes and then taken up by cells. Trying to keep the area sterile by typical biology techniques would also compromise your storage DNA. Oh, and also if you put it into an organism, you better pick a plant. Many organisms (particularly microbes) are under strong selective pressure to excise DNA that isn't essential or very useful. Even in plants, your sequence is going to get corrupted over time by mutations down the generations (or even just in the one generation if your plant lives a very long time).
nicolashahn 8y ago
Hold on. So we now have the ability to create DNA with arbitrary code, and we also have the ability to sequence a genome. Does this mean we're able to clone a human from just their DNA encoding stored digitally? Like, I could swab some DNA, store it on some format that will persist, and my descendents can download and print their great great grandfather in 100 years?

51 points by jkuria 8y ago

This reminds me of "The Chase" from Star Trek: The Next Generation. Its plot centred around data stored in the DNA of different species across the galaxy.

presscast 8y ago

Isn't DNA orders of magnitude slower than current storage media?

jfarlow 8y ago
Yes. Many orders of magnitude. However, if you're just doing cold storage, the data-density of DNA is many orders of magnitude greater. With essentially sub-nanometer 3D atomic data-density you can pack a staggering amount of data in a very very small package. And though we're not quite there yet in read/write speeds at the ends of the route, the old saying rings true about the fastest way to transfer large amounts of data between the coasts - in a station-wagon full of hard drives.
- Real_S 8y ago
  Also, consider that read-speed is rapidly improving. So for long-term storage, reading the data at a reasonable rate in the future is almost guaranteed. Hopefully write-speed will reach a similar rate of improvement. Importantly, expected corruption of the data stored in DNA is less than what one expects using other forms of data storage. So if stability of the data is a concern, DNA has unique advantages.
  - drvdevd 8y ago
    I wonder what effect the introduction of artificial nucleotides and engineered proteins will have on “DNA” storage speeds as well...

technobabble 8y ago

Does anybody recommend specific research papers on how this is done?

jfarlow 8y ago
- High level article [1] - A "Roadmap" from the semiconductor/data side, looking at DNA from its first principles [2] - Review combining the tech side with the bio side [3] [1] http://www.sciencemag.org/news/2017/03/dna-could-store-all-worlds-data-one-room http://www.sciencemag.org/news/2017/03/dna-could-store-all-w... [2] https://www.src.org/program/grc/semisynbio/ https://www.src.org/program/grc/semisynbio/ [see PDF on right side of page] [3] https://www.biorxiv.org/content/biorxiv/early/2017/03/07/114553.full.pdf https://www.biorxiv.org/content/biorxiv/early/2017/03/07/114...

onychomys 8y ago

It's super slow, hard to read, easily corruptible, and expensive. Literally the only advantage is storage density. But we live in a world where giant companies build immense data centers in places nobody cares about, like Nevada. So I'm not sure if anybody is actually worried about storage density.

i_am_nomad 8y ago
Can we maybe cut the coastal snobbery just a bit? Plenty of people care about Nevada, specifically, Nevadans. (Not sure that’s the correct demonym.)
- choward 8y ago
  Thanks. It is the correct demonym. Just because there is a lot of unused land doesn't mean nobody cares.
PyroLagus 8y ago
It maybe easily corruptible, but it's super easy to copy. Nature has already solved that part for us already. If you can just have some bacteria make a million copies of your data, it doesn't matter if even several thousands of the strands become corrupted. But of course, unless you can actually read several hundred thousand strands in a reasonable amount of time, that redundancy doesn't help too much. I wonder if we could engineer bacteria to facilitate reading and writing to DNA in a more efficient manner.
MauranKilom 8y ago
Let's not forget one church building archiving the internet several times over.

calgoo 8y ago

"Someone creates a virus to delete or encode all your data that's stored in DNA secure storage. This then spreads and starts to encrypt the DNA sequence in a humans body. A new step in computer malware happens where malware writers start modifying people and unless you pay the 100 bitcoin to unlock your DNA in 24 hours you are dead." Its an interesting time to live in. On the one hand i love the push for all the new tech we are creating. However, at the same time, the old problems dont go away because we create new tools.

Real_S 8y ago
Interesting thoughts. Encrypting dna molecules is possible, but it probably wouldn't use a virus. Also, if your cell's DNA was changed like this, it would kill you very quickly, interfering with normal cellular function.
- jobigoud 8y ago
  A better strategy would be to target the gametes and encrypt these only. Now if you want to be unsterilized you have to pay. Maybe it could a contraceptive method.
mr_toad 8y ago
This makes about as much sense as a computer virus that destroyed the entire OS of a machine it infected. It would never be able to propagate.
- staplers 8y ago
  Unless it propagates before the trigger is pulled
  - thomasfedb 8y ago
    Lots of examples of this in extant biology - kill genes and mechanisms for destroying cells and DNA that are vital to the survival of multicellular organisms such as humans.

Real_S 8y ago

Also consider that cryptography can be applied directly to DNA molecules. Cryptography for genetic material: https://www.biorxiv.org/content/early/2017/06/30/157685 https://www.biorxiv.org/content/early/2017/06/30/157685

p1mrx 8y ago
Why would a new storage medium require a new method of cryptography? These problems seem orthogonal.
- emmelaich 8y ago
  All algorithms have engineering trade-offs to for practical application. e.g. available memory, cpu, time. DNA is very slow but very dense.

pronoiac 8y ago

"Massive Attack’s Mezzanine “remastered” in aerosol spray can format," as DNA > The complex encoding process is explained in detail in by Dr Robert Grass of Zurich-based company TurboBeads. “This digital bitstream of the album (0s and 1s) was first translated to 901’065 DNA sequences (A, C, T and Gs), each 105 characters long”, says Grass. “The 901’065 individual sequences were then chemically synthesised resulting in a synthetic DNA sample, which fully represents the digital bitstream of the album.” > Then, in order to “guarantee information stability”, the DNA sequences were encapsulated in “synthetic glass fossils”, which were added directly to the spray can. According to Dr Grass, each can “contains at least 0.1 micrograms of the synthetic DNA, which is equivalent to 1 million copies of the album.” Playback is left as an exercise for the listener http://www.factmag.com/2018/10/19/massive-attacks-mezzanine-remastered-spray-can/ http://www.factmag.com/2018/10/19/massive-attacks-mezzanine-...

deepnotderp 8y ago

Last I looked into this (I think micron was looking into it also?) these were the major problems: 1. Orders of magnitude slower than normal memory. 2. Could be easily damaged relative to current technology. 3. Requires orders of magnitude more energy than standard storage media. 3. Is currently far more expensive than standard storage media as well. And unfortunately, due to the fundamental nature of the storage media, I don't know if the orders of magnitudes jumps that we're used to with conventional storage/ memory technology will be possible here. So the only real advantage is density, but if it's going to be used as cold storage anyways, does anyone really care compared to tape? And is this ultimately more promising than atoms on a surface (e.g. like IBM millipede)?

pcrh 8y ago
It also has the potential to be introduced into a living organism. In this manner it could well last for millennia by being transmitted to the organism's descendants. Assuming there is no disadvantage of carrying this DNA it could perdure over geological periods of time.
- colejohnson66 8y ago
  Until mutations happen. But redundancy should help with that
wuschel 8y ago
I had similar results when looking into this - it was not competitive in cost or performance (speed, errors) a year ago.

Apocryphon 8y ago

If we're going to talk about alternative computing media, how's optical circuitry doing these days?

ImaCake 8y ago

What often seems to be overlooked when talking about using DNA for storage is that microorganisms want to eat it for their own uses. DNA in the environment isn't just ignored, it is quickly degraded by excreted enzymes and then taken up by cells. Trying to keep the area sterile by typical biology techniques would also compromise your storage DNA. Oh, and also if you put it into an organism, you better pick a plant. Many organisms (particularly microbes) are under strong selective pressure to excise DNA that isn't essential or very useful. Even in plants, your sequence is going to get corrupted over time by mutations down the generations (or even just in the one generation if your plant lives a very long time).

nicolashahn 8y ago

Hold on. So we now have the ability to create DNA with arbitrary code, and we also have the ability to sequence a genome. Does this mean we're able to clone a human from just their DNA encoding stored digitally? Like, I could swab some DNA, store it on some format that will persist, and my descendents can download and print their great great grandfather in 100 years?