Breed-Solomon
July 30, 2015 8:58 AM   Subscribe

Since it folds in three dimensions, we could store all of the world’s current data—everyone’s photos, every Facebook status update, all of Wikipedia, everything—using less than an ounce of DNA. And, with its propensity to replicate given the right conditions, millions of copies of DNA can be made in the lab in just a few hours. Such favorable traits make DNA an ideal candidate for storing lots of informations, for a long time, in a small space.
But how stable is DNA? The Reed-Solomon method, long used to error-check data transmission and duplication, is now being explored as an adjunct to the long-term archiving of information encoded in DNA. A post by Alex Riley at the PBS Science blog NOVA/NEXT.
posted by Rumple (33 comments total) 20 users marked this as a favorite
 
But how stable is DNA? The Reed-Solomon method, long used to error-check data transmission and duplication

Presumably the millions of copies would provide for robust error-checking, no?
posted by leotrotsky at 9:04 AM on July 30, 2015 [1 favorite]


Does this mean that we are closer to the day where I can store all of my cat .gifs on my cat?
posted by GenjiandProust at 9:06 AM on July 30, 2015 [15 favorites]


The Angew. Chemie paper by Robert Grass et. al..
posted by James Scott-Brown at 9:10 AM on July 30, 2015 [1 favorite]


Given that is the case, it really begs the question if awareness, emotions, and consciousness couldn't exist fully formed as well or better than a humans in a minute scale that we have no capacity of understanding. How do we know there aren't bacteria or single celled organisms who aren't in fact emotionally and intellectually more competent than we are? When we think about finding life on other planets we are willing to imagine life as radically different that how we define it on earth, but are we not willing to acknowledge we may be missing things going on right in front of us?
posted by xarnop at 9:24 AM on July 30, 2015


What's more why do we assume that the most precious aspect of life itself should be used as a tool without regard for it's experience of being? Isn't it possible there is a reason why all living beings are set to PROTECT AND REPAIR THE DNA if at all possible? Perhaps a precious part of our life force lives there.

Forgive us, we know not what we do.
posted by xarnop at 9:36 AM on July 30, 2015


xarnop: I know you are coming from a place of care, but vitalism has been discredited for hundreds of years. There have been literally thousands of experiments done to test for this idea, all coming up negative. Inanimate matter such as DNA does not have consciousness.

Even if it were true, living creatures cannot function without massive continual destruction of DNA.

Every biological source of calories - which is to say every living thing we eat - has DNA which is destroyed in said extraction of energy and nutrients.

Inside us, as we speak, quintillions of pieces of DNA every second are being violently chopped reorganized, recycled and destroyed up by our own biological machinery. It happens in every living being. It (and it's cousin RNA) is literally both the fuel and instruction manual of life's tapestry and without it's continual destruction and reorganization life as we know it would not exist.

If we are willing to eat an organic lettuce for a nice flavor (destroying DNA) - or making pigments from plants (destroying DNA) - we should be willing to taking some synthetic lab-made DNA and re-organizing its nucleotide sequence for the the preservation of the long-term cultural interests of humanity.
posted by lalochezia at 9:50 AM on July 30, 2015 [20 favorites]


What if the spirit moves us through the pituitary gland

Perhaps hard drives are sentient
posted by benzenedream at 10:06 AM on July 30, 2015 [3 favorites]


How do we know there aren't bacteria or single celled organisms who aren't in fact emotionally and intellectually more competent than we are?

I sometimes wonder about how many tiny insects and what-have-you I kill walking through a field, so the idea of a world where each course of antibiotics or bolus of bleach down the bog is equivalent to genocide is fairly horrifying.
posted by sobarel at 10:06 AM on July 30, 2015 [1 favorite]


xarnop: How do we know there aren't bacteria or single celled organisms who aren't in fact emotionally and intellectually more competent than we are?

You may be interested in the work of the biophysicist Eshel Ben-Jacob on the collective intelligence of bacteria. (I learned just now that he sadly passed away last month.) This sounds less fringe-y that it actually is; he was highly esteemed in the scientific community and e.g. former president of the Israel Physical Society.
posted by tecg at 10:14 AM on July 30, 2015 [1 favorite]


I forgot to add that Ben-Jacob never suggested that individual bacteria possess consciousness, but he cam close to suggested something like that for populations of bacteria.
posted by tecg at 10:17 AM on July 30, 2015


place these glass beads in the dark at –0.4˚ F, the conditions of the Svalbard Global Seed Bank on the Norwegian island of Spitsbergen, and you could save your photos, music, and eBooks for two million years.
I'm still thinking about this.
posted by doctornemo at 10:26 AM on July 30, 2015


I am aware of research into bacterial intelligence as well as slime molds and quorum sensing. Just because I'm an animist doesn't mean I don't read science.

It just means I cringe in horror at how unwilling humans are to consider that other forms of sentience might exist. Reading up on elephant intelligence, despite that they appear to mourn their dead, exhibit empathy for other elephants AND non elephants we find that "Scientists often debate the extent that elephants feel emotion." This kind of crap is why I don't trust "ethics committees" of scientists to actually favor the possibility of sensing or emotion over the pressure to plow ahead and presume there is no feeling if at all possible or if there is any doubt of certainty there is emotion/sensing present.

I do think it's possibly other forms of energy than just life can sense, in fact I wonder if life isn't the manifestation of an already sensing universe fighting as hard as it can to get to the locations and experiences it desires despite the difficulty of exerting any will against either the order or the chaos defined by the physics of this world.
posted by xarnop at 10:48 AM on July 30, 2015


Speaking of elephants, from the same blog
posted by Rumple at 11:04 AM on July 30, 2015


It just means I cringe in horror at how unwilling humans are to consider that other forms of sentience might exist.

Nonsense. Humans have always been more than willing to believe that other forms of sentience exist. In fact many of them orient their lives around that belief, and what they think those sentiences want. They've sacrificed goods and animal and people to those sentiences. They've slaughtered people of differing ethnic and cultural backgrounds in the name of these sentiences. And currently, evidently these sentiences are obsessed with the sexual morality of humans.

So what you ought to be asking is not "why don't people believe?", but "How can I make money and power off of this exploit in human cognition?" Play your cards right, and you can probably get followers who will enact a crusade against people who don't believe in Our Sentient DNA buddies, or the Will of the Cosmos, or whatever.
posted by happyroach at 11:19 AM on July 30, 2015 [5 favorites]


There are lots of good ways of storing huge amounts of data, if you don't worry about trivialities like storing or retrieving it in a reasonable amount of time, doing a random seek, or the physical robustness of it all.
posted by Mitrovarr at 11:20 AM on July 30, 2015 [5 favorites]


xarnop: Isn't it possible there is a reason why all living beings are set to PROTECT AND REPAIR THE DNA if at all possible? Perhaps a precious part of our life force lives there.

Actually, they aren't. Living things will generally destroy any DNA found outside a cell. It's an important safeguard against viruses. It'll also happily sacrifice cells for the good of the whole, destroying the DNA inside, as part of apoptosis. This happens to huge numbers of cells all the time.

Anyways, there are obvious good reasons to protect the DNA in a living cell from damage. It's essential for life, and cells that mutate can cause cancer, etc...
posted by Mitrovarr at 12:07 PM on July 30, 2015 [2 favorites]


*reads thread, shrugs, returns to the bench to create and destroy DNA again and again and again and again*

PCR, DNase, PCR, DNase, PCR, DNase, bleach bleach bleach.
posted by maryr at 12:34 PM on July 30, 2015 [3 favorites]


Given that is the case, it really begs the question if awareness, emotions, and consciousness couldn't exist fully formed as well or better than a humans in a minute scale that we have no capacity of understanding.

You know, I'm relatively well-disposed, in a purely speculative way, to the notion that on some level consciousness is an inextricable attribute of matter. But I've always thought the obvious conclusion is that there's not so much special about "life" versus "not life" after all, and little significance on a grand scale to the continuity of any particular "experience of being."
posted by atoxyl at 12:44 PM on July 30, 2015


Anyway this is really cool. Definitely strictly an archival format for now though.
posted by atoxyl at 12:51 PM on July 30, 2015


Someone who knows Reed-Solomon well: Is it localized enough for the possibility of a single protein (or ribonucleoprotein) to use the correction codes during replication? I.e. are the "checksums" (or whatever is used in Reed Solomon) placed right next to the data, so that a modified DNA polymerase could use, say, every 4th base to check on the accuracy of the previous 3?
posted by clawsoon at 2:06 PM on July 30, 2015


> Does this mean that we are closer to the day where I can store all of my cat .gifs on my cat?

22 million years in the future, whiskered scientists jostle around a display. "Wait, that section that keeps repeati-"
"It's noncoding junk, full of random mutations."
"Just, bring it up again. Reverse it. Okay, try losing this outside part?"
"This is a waste of time."
"Wait, run it in parallel lines instead."
As one, their heads slowly tilt to the side. "My god, it — it looks like a face!"
"It looks like a message, and it looks... grumpy."
posted by lucidium at 3:00 PM on July 30, 2015 [3 favorites]


Someone who knows Reed-Solomon well: Is it localized enough for the possibility of a single protein (or ribonucleoprotein) to use the correction codes during replication? I.e. are the "checksums" (or whatever is used in Reed Solomon) placed right next to the data, so that a modified DNA polymerase could use, say, every 4th base to check on the accuracy of the previous 3?

It's not something I know well (it's something I learned once upon a time in CS school) but I think it would depend on the exact scheme chosen? A message N symbols long (a symbol could be more than one DNA base if you wanted) is encoded in blocks of K symbols and can withstand the corruption of (K-N)/2, which is to say you do need to be able to look at the other (K+N)/2.
posted by atoxyl at 3:18 PM on July 30, 2015


Note that it's not a checksum and there are not particular "redundancy" data points. The beauty of it is that any combination of (K+N)/2 symbols from a block is sufficient to reconstruct the message.
posted by atoxyl at 3:27 PM on July 30, 2015


N being the number of message symbols encoded per block, not the total length of everything you are encoding. That's probably clear from context but my wording was poor.
posted by atoxyl at 3:34 PM on July 30, 2015


Given that is the case, it really begs the question if awareness, emotions, and consciousness couldn't exist fully formed as well or better than a humans in a minute scale that we have no capacity of understanding. How do we know there aren't bacteria or single celled organisms who aren't in fact emotionally and intellectually more competent than we are?

This is fun and you should read it
posted by a snickering nuthatch at 7:40 PM on July 30, 2015


Is it possible to read back DNA-stored data in a fast, cost-effective way yet? If not, magnetic and optical storage will still be king and queen.
posted by ymgve at 8:46 PM on July 30, 2015


so that a modified DNA polymerase could use, say, every 4th base to check on the accuracy of the previous 3?

This is kind of the opposite of cousin wobble.
posted by maryr at 9:48 PM on July 30, 2015


Actually, they aren't. Living things will generally destroy any DNA found outside a cell.

There are also organisms that regularly jettison large parts of their genome (the one I'm thinking of actually also has two nuclei and over 15,000 "nanochromosomes"...Oxytricha, never change, you little weirdo!).
posted by en forme de poire at 12:16 AM on July 31, 2015


(Also most papers about bacterial "intelligence" or information processing that I'm aware of have focused on properties of populations of bacteria, not on single individuals. But that's kind of a derail here.)
posted by en forme de poire at 12:19 AM on July 31, 2015


Ben-Jacob never suggested that individual bacteria possess consciousness, but he came close to suggested something like that for populations of bacteria

Speaking as a population of bacteria, I endorse this notion.
posted by flabdablet at 4:27 AM on July 31, 2015 [3 favorites]


As one, their heads slowly tilt to the side. "My god, it — it looks like a face!"
"It looks like a message, and it looks... grumpy."


http://www.youtube.com/watch?v=wSYAZnQmffg
posted by flabdablet at 4:36 AM on July 31, 2015


Is it possible to read back DNA-stored data in a fast, cost-effective way yet?

As I understand it, full-genome DNA sequencing gets done by massively replicating it, then breaking it into pieces that are short enough to sequence directly, sequencing those, then working out where the breaks were by matching end pieces to the middle parts of other pieces.

If I were to design a digital storage system that relied on this kind of thing, I would build my archival molecules as data-bearing blocks with embedded error correction codes, separated by cleavage regions, using a group code recording scheme inside the data blocks so that data can never look like a cleavage marker; then I'd store all the data in the form of Merkle trees, allowing for easy data reassembly after a massively parallel sequencing step.

With enough parallelism in that sequencing step, it ought to be possible to make large reads very fast.
posted by flabdablet at 5:00 AM on July 31, 2015 [1 favorite]


There are also newer technologies that give much longer but somewhat more error-prone reads (this describes PacBio, at least last I checked -- they may have gotten the error rate down since then) but they are not as common yet compared to e.g. Illumina-style 100bp reads.
posted by en forme de poire at 2:55 PM on August 1, 2015 [1 favorite]


« Older Why You Should Never Say: ‘Beauty Lies in the Eye...   |   “Writing is healing. Writing is art. Writing is... Newer »


This thread has been archived and is closed to new comments