How can you write data to DNA without changing the base sequence?
- Thread starter JournalBot
- Start date
Brings a whole new meaning to Genetic Algorithms...
.... I'll see myself out.
But in all seriousness, the fact that we're still working on this tech is neat. Because for large datasets, it's important that we continue to develop more data-dense storage media. The thing that the article doesn't get into, that I'd like to know, what's the read/write speed look like for this? The how is there, and the fact that the reliability is still... a way's off, but is this one of those slower WORM(write-once-read-many)-type situations?
.... I'll see myself out.
But in all seriousness, the fact that we're still working on this tech is neat. Because for large datasets, it's important that we continue to develop more data-dense storage media. The thing that the article doesn't get into, that I'd like to know, what's the read/write speed look like for this? The how is there, and the fact that the reliability is still... a way's off, but is this one of those slower WORM(write-once-read-many)-type situations?
Upvote
24
(24
/
0)
Upvote
30
(30
/
0)
It's a typo. The paper states 95°C. DNA denatures at 100°C.
Last edited:
Upvote
25
(25
/
0)
Out of curiosity, will your cat violate copyright law by having kittens? The fact that the videos are stored in epigenetic data suggests that reproduction won't be a DCMA violation, but it would be good to have a biologist-lawyer confirm that hunch.
Upvote
31
(31
/
0)
Upvote
-7
(7
/
-14)
CRISPR genetic splicing. You can also use key sequences. For example lets say you have a AGGAAGGGAAAGGG A = 1 G = 0. You have created a binary sequence, and if you have the appropriate RNA you can bind to that specific sequence of 10011000111000, or more appropriately the strand behind it.
Sequencing no longer requires the entire strand be built up or broken down. It's macro building now.
The problem with genetic splicing is that proteins tends to be very unpredictable sometimes and they randomly die and mutate due to environmental conditions.
Sequencing no longer requires the entire strand be built up or broken down. It's macro building now.
The problem with genetic splicing is that proteins tends to be very unpredictable sometimes and they randomly die and mutate due to environmental conditions.
Upvote
5
(6
/
-1)
DNA tends to be referred to as write-once, read-never. More accurate would probably be "read rarely". It is anticipated that the best use-case for such storage would be as backup to more traditional "live" media and/or for stuff with a high-likelihood will simply never need to be accessed again.
Upvote
-3
(4
/
-7)
Upvote
8
(10
/
-2)
Thank goodness we have really smart and inquisitive people working in science and other fields whose efforts sometimes lead to breakthroughs the rest of us benefit from, or simply provide us with a better life. Knowledge rocks!
Upvote
12
(12
/
0)
Sorry in advance for being pedantic.
"Epigenetics" refers to heritable modifications that affect gene expression. Epigenenic modifications call also be made to proteins and other cell components (some of the best characterized are histone modifications). DNA base modification, by itself, is not necessarily 'epigenetics' because not all modifications persist to the next generation or even through the next cell division. So, some DNA base modifications are epigenetic while others aren't and it's hard to tell which is the case without tracking the changes over time.
And really, epigenetics has largely turned into a buzzword or marketing term for grants in research to try to differentiate the study of the underlying mechanics from other, previously identified regulatory mechanisms. Kinda like how the suffix 'omics' gets added to research field names. At the core, epigenetics includes a number of interesting mechanisms within the larger context of biological regulation.
"Epigenetics" refers to heritable modifications that affect gene expression. Epigenenic modifications call also be made to proteins and other cell components (some of the best characterized are histone modifications). DNA base modification, by itself, is not necessarily 'epigenetics' because not all modifications persist to the next generation or even through the next cell division. So, some DNA base modifications are epigenetic while others aren't and it's hard to tell which is the case without tracking the changes over time.
And really, epigenetics has largely turned into a buzzword or marketing term for grants in research to try to differentiate the study of the underlying mechanics from other, previously identified regulatory mechanisms. Kinda like how the suffix 'omics' gets added to research field names. At the core, epigenetics includes a number of interesting mechanisms within the larger context of biological regulation.
Last edited:
Upvote
10
(10
/
0)
This could be a good thing but knowing humanity there is someone who will think that they are 'God' before long and if we don't bow before them . . . Who knows what will happen next!
Upvote
-10
(0
/
-10)
Where ever did you get that idea from? The temperature is dependent mostly on the sequence, conformation, and ion concentration. Also, only the hardiest of DNA sequences (certain telomeric-like sequences) manage to denature at approx 95C. The temperature in the paper is not the denaturing temperature. They need to denature the DNA to read it back, ie. get the single strands through the pores.
Upvote
8
(8
/
0)
Great. Now just improve the read/write speeds & cost per bit about a billion fold & they’ll be laughing. Instead of me.
Upvote
-6
(0
/
-6)
A subtle but crucial point that may be missed in this article is that the "DNA template" (the paper calls it the "DNA carrier") being "printed" onto must be specially synthesised so that each "DNA brick" (paper calls it "DNA movable type") can only bind to it at one and precisely one location. Therefore, this technique cannot be used to epigenetically store data by piggybacking on arbitrary DNA sequences.
For example, in order to store some arbitrary sequence of 8 bits, assuming that each brick encodes one bit, you need to have 8 unique binding sites on the DNA template, and you need to have 16 different DNA bricks at your disposal (two for each site, representing binary "1" and "0" respectively). This scales linearly, so N bits of storage requires 2N unique bricks.
This further implies that as N grows, each brick would have to grow longer (in terms of underlying base pairs) to guarantee that it binds to one and precisely one point on the template*. In other words, this technique requires more than one base pair to encode a single bit of information. For comparison, a direct-sequence-synthesis scheme can encode two bits of information per base pair. Therefore, this technique also will have lower areal (linear?) information density that a direct-sequence-synthesis method.
* There are 4^(m-1)*3 potential sequences for a length-m DNA strand, assuming that palindromes are disallowed, but wait, the CpG sequence has to occur precisely once at or near the middle and nowhere else...exercise for the reader!
For example, in order to store some arbitrary sequence of 8 bits, assuming that each brick encodes one bit, you need to have 8 unique binding sites on the DNA template, and you need to have 16 different DNA bricks at your disposal (two for each site, representing binary "1" and "0" respectively). This scales linearly, so N bits of storage requires 2N unique bricks.
This further implies that as N grows, each brick would have to grow longer (in terms of underlying base pairs) to guarantee that it binds to one and precisely one point on the template*. In other words, this technique requires more than one base pair to encode a single bit of information. For comparison, a direct-sequence-synthesis scheme can encode two bits of information per base pair. Therefore, this technique also will have lower areal (linear?) information density that a direct-sequence-synthesis method.
* There are 4^(m-1)*3 potential sequences for a length-m DNA strand, assuming that palindromes are disallowed, but wait, the CpG sequence has to occur precisely once at or near the middle and nowhere else...exercise for the reader!
Last edited:
Upvote
8
(8
/
0)
Embarrassing to use a graphic of left handed DNA for a story that is about DNA. Some of us just can't un-see it
Upvote
1
(1
/
0)
Tbh i hadn't realised dna was stable as anything like those temperatures. Way to go, dna!
Upvote
1
(1
/
0)
This is about storage. You need neurons to run Doom. Like this
View: https://m.youtube.com/watch?v=bEXefdbQDjw
Upvote
4
(4
/
0)
I love the (admittedly silly) idea that we are just reverse engineering. Amusing to thing our biosphere is just long term data storage that was abandoned when it became corrupted and began self modifying.
Upvote
1
(1
/
0)
Upvote
-6
(0
/
-6)
Upvote
-3
(0
/
-3)
DNA does have proven scalability, of course. Even humans have a very affordable steady state sequence/watt ratio and net throughput.
Upvote
0
(0
/
0)
I'm sure everyone here in the Ars comment section either LOLed or smacked their forehead.
Upvote
1
(1
/
0)
Information is more useful than snark, and the obvious isn't obvious to everyone.
Upvote
0
(0
/
0)
I have a theory that left handed DNA graphics generate comments about their left handedness, and this boosts their weight in AI training sets. If true, pretty soon, all AI generated DNA graphics will be left handed.
Upvote
1
(1
/
0)