3 ms·

Beyond the Hard Drive: Encoding Data in DNA

4 points by dsr12 8y ago

dekhn 8y ago
Although I am a big fan of encoding information in DNA, nothing in the arguments for it make sense, economically speaking, compared to more mature industries like hard drives and flash. People often cite the dramatic growth of storage as a motivation for DNA, but realistically, the amount of storage produced each year is driven by the demand of the previous year + some expected growth. If there was a economic case for needing more storage, vendors would build plants and produce more storage and invest in making mature technologies more efficient. The claims about DNA being stable are only true under very limited conditions. It would be a significant challenge to store large amounts of DNA reliably, compared to tape or flash.
- ryanmercer 8y ago
  >Although I am a big fan of encoding information in DNA, nothing in the arguments for it make sense, economically speaking, compared to more mature industries like hard drives and flash. The only application I could ever think that would ever remotely make sense would be when genetically engineering things you could use the 'junk DNA' to put information in about what you designed however, with enough time/generations, it would be garbled.
  - dekhn 8y ago
    it would be a terrible idea to put data into the junk DNA sections. They're highly variable, under weird selection, and may even have a function that, when replaced with other DNA, could have a deleterious effect. That said I do think you could replace large swaths of LINES and SINES and Alu without any negative functionality. I think Reed Solomon encoding should be quite robust.
ryanmercer 8y ago
I wrote about this back in 2013 comparing what we were able to encode in DNA at the time with what nature can :) https://www.ryanmercer.com/ryansthoughts/2013/3/28/dna-data-storage.html https://www.ryanmercer.com/ryansthoughts/2013/3/28/dna-data-... It was a bit of a task trying to figure out how much a strand of human DNA weighed (turns out about 0.000000001 picograms).

4 points by dsr12 8y ago

Although I am a big fan of encoding information in DNA, nothing in the arguments for it make sense, economically speaking, compared to more mature industries like hard drives and flash. People often cite the dramatic growth of storage as a motivation for DNA, but realistically, the amount of storage produced each year is driven by the demand of the previous year + some expected growth. If there was a economic case for needing more storage, vendors would build plants and produce more storage and invest in making mature technologies more efficient. The claims about DNA being stable are only true under very limited conditions. It would be a significant challenge to store large amounts of DNA reliably, compared to tape or flash.

ryanmercer 8y ago
>Although I am a big fan of encoding information in DNA, nothing in the arguments for it make sense, economically speaking, compared to more mature industries like hard drives and flash. The only application I could ever think that would ever remotely make sense would be when genetically engineering things you could use the 'junk DNA' to put information in about what you designed however, with enough time/generations, it would be garbled.
- dekhn 8y ago
  it would be a terrible idea to put data into the junk DNA sections. They're highly variable, under weird selection, and may even have a function that, when replaced with other DNA, could have a deleterious effect. That said I do think you could replace large swaths of LINES and SINES and Alu without any negative functionality. I think Reed Solomon encoding should be quite robust.

ryanmercer 8y ago

I wrote about this back in 2013 comparing what we were able to encode in DNA at the time with what nature can :) https://www.ryanmercer.com/ryansthoughts/2013/3/28/dna-data-storage.html https://www.ryanmercer.com/ryansthoughts/2013/3/28/dna-data-... It was a bit of a task trying to figure out how much a strand of human DNA weighed (turns out about 0.000000001 picograms).