3 ms·

Researchers Store 90GB of Data in 1g of Bacteria

9 points by dfield 16y ago

gregable 16y ago
They stored a few kb of data replicated millions of times. That's much different.
wgrover 16y ago
"an expensive sequencer is needed to retrieve data" Given how long it still takes to sequence an individual's genome ("only" 3 gigabases), this may not be practical any time soon. Anyway, generating astronomical amounts of data really small using DNA isn't that hard - say you synthesize short pieces of DNA, each 34 bases long, randomly selecting A, C, T or G for each base (easy to do with a commercial DNA synthesizer). If you make about a millimole (10^20 molecules) of this DNA, odds are that each piece of DNA has a unique sequence. You could have your own little jar containing 10^20 unique molecules. If the presence of a particular piece of DNA represents a TRUE for a given bit, you have about 10 exabytes of data storage in a jar. Now, operating on that data is another story...
- bhickey 16y ago
  It doesn't take particularly long to sequence a human genome. If you want long reads, you need to spend more time sitting on the machine. Last I checked, the limiting steps were contention for machine time and library prep. If you're doing a custom library prep it can be very time consuming, but I believe the trend is moving toward whole exome sequencing. We should see sub-$5k exomes (at scale) in 2011.
  - wgrover 16y ago
    You're definitely right. I just meant that, until you can sequence a genome faster and cheaper than you can read 90 GB off of a hard drive, storing information in DNA and reading it out via sequencing may not be feasible.
vehementi 16y ago
They didn't actually store it, they "showed how to encrypt it" and extrapolated that they should be able to store that much...
- bhickey 16y ago
  Spot on. The article is garbage. The Declaration of Independence is about 8000 bytes, without parity you could encode the whole thing in 378 bases. With a little more care you could avoid generating open reading frames (ensure that the sequence 'ATG' never appears). There is nothing novel or exciting about storing this amount of data in bacteria. If you grow multiple copies of it, you haven't stored 90gb of data, you've simply made copies of your original 8k.
ghshephard 16y ago
I'm looking forward to the follow-up post where Researchers restore 90 GB of Data from 1g of Bacteria.

9 points by dfield 16y ago

They stored a few kb of data replicated millions of times. That's much different.

"an expensive sequencer is needed to retrieve data" Given how long it still takes to sequence an individual's genome ("only" 3 gigabases), this may not be practical any time soon. Anyway, generating astronomical amounts of data really small using DNA isn't that hard - say you synthesize short pieces of DNA, each 34 bases long, randomly selecting A, C, T or G for each base (easy to do with a commercial DNA synthesizer). If you make about a millimole (10^20 molecules) of this DNA, odds are that each piece of DNA has a unique sequence. You could have your own little jar containing 10^20 unique molecules. If the presence of a particular piece of DNA represents a TRUE for a given bit, you have about 10 exabytes of data storage in a jar. Now, operating on that data is another story...

bhickey 16y ago
It doesn't take particularly long to sequence a human genome. If you want long reads, you need to spend more time sitting on the machine. Last I checked, the limiting steps were contention for machine time and library prep. If you're doing a custom library prep it can be very time consuming, but I believe the trend is moving toward whole exome sequencing. We should see sub-$5k exomes (at scale) in 2011.
- wgrover 16y ago
  You're definitely right. I just meant that, until you can sequence a genome faster and cheaper than you can read 90 GB off of a hard drive, storing information in DNA and reading it out via sequencing may not be feasible.

vehementi 16y ago

They didn't actually store it, they "showed how to encrypt it" and extrapolated that they should be able to store that much...

bhickey 16y ago
Spot on. The article is garbage. The Declaration of Independence is about 8000 bytes, without parity you could encode the whole thing in 378 bases. With a little more care you could avoid generating open reading frames (ensure that the sequence 'ATG' never appears). There is nothing novel or exciting about storing this amount of data in bacteria. If you grow multiple copies of it, you haven't stored 90gb of data, you've simply made copies of your original 8k.

ghshephard 16y ago

I'm looking forward to the follow-up post where Researchers restore 90 GB of Data from 1g of Bacteria.