Microsoft reports they are encoding large blocks of digital data (even music videos) into DNA ... does anyone know more the exact method they are using and have an opinion on whether it may ever be made an efficient/economical alternative to current storage technologies?
"DNA is an amazing information storage molecule that encodes data about how a living system works. We're repurposing that capacity to store digital data - pictures, videos, documents ... This is one important example of the potential of borrowing from nature to build better computer systems." ~~Luis Henrique Ceze
http://www.msn.com/en-us/video/other/microsoft-successfully-stored-ok-go-music-video-on-dna-strands/vp-BBu7bky
Exactly how is digital data encoded into DNA?
As I understand it and to be clear, there is no question here of encoding digital data into existing DNA, e.g. of a bacteria, cell, …..(as one could think at first quick reading). Neither is it intended to use the elements of the cell machinery like RNA transcription, etc.
The idea here is to produce synthetic DNA (which encodes digital information) using an oligonucleotide synthesis machine. The information can then be retrieved at will using a DNA sequencing machine.
Why DNA and not another synthetic chemical molecule ? Because, besides its major target (storing information for life and reproduction) nature has endowed DNA with 3 very interesting “lateral” features (also of high necessity for a good storage of information for life and reproduction) :
- Capacity to wind up into an extremely compact volume : each one of our cells contains a DNA molecule with 3.109 bp (base pairs).
- Capacity to hold its integrity during a very long period (in cold, dry and dark conditions) : see DNA of extinct species.
- Capacity to be copied many times (nearly) without loss of information : this is vital for living species and inherent to its double-helix structure (so with the PCR – polymerase chain reaction - method, it is possible to multiply DNA strands ad libitum).
However, using DNA to store digital data has a drawback : retrieving the information with a sequencing machine must be performed base after base. It takes time. But technology development will most probably reduce this drawback. And it is also not bad to ascertain an high degree of reliability of the retrieved data. To be sure, several (4) copies of the strands can also be systematically made at encoding stage. This would allow error free storage for millions of years.
So synthetic DNA is used for its outstanding storage and reliability features.
As there is no question of storing the information into bacteria, cells, …., devices like graphene nanopore devices are developed to allow the storage of the produced synthetic DNA, as is done with books in a library. But as mentioned in a post above, the manufactured DNA-molecules can also be put in a test tube and dehydrated.
How the data are encoded ? As already mentioned in other posts, a correspondence is made between the Boolean « 0 » and « 1 » and the 4 bases A, T, C, G (= Adenine, Thymine, Cytosine and Guanine) of DNA.
For instance : T and G are given the value « 1 » and A and C the value « 0 ». So the digital information « 01100110 » would be encoded by letting the oligonucleotide synthesis machine synthesize the oligonucleotide « ATGAATTC ». This oligonucleotide can then be read by a DNA sequencing machine to restore the initial information « 01100110 ». Note that there also exists a possibility of base 3 (« 0 », « 1 », « 2 ») encoding.
What I don’t understand is why to give arbitrarily the value « 1 » to two bases and the value « 0 » to the two others no matter how ? Of course, once the choice made, the information will be univocally stored.
However, one could do it simpler : just give the value « 1 » to G (or C) and the value « 0 » to A (or T) or vice-versa.
Indeed, one wants to store DNA which is a double-stranded helix to use its compact storage capacity. So, why not use a basic characteristic of DNA : the fact that the 4 bases are coupled 2 by 2. G is (chemically) coupled with C and A with T. This is the essence of creating a copy of DNA at cell division. After the two strands are being separated, DNA polymerase adds the complementary bases to the existing bases of the original DNA. For instance (capitals : one helix ; lowercases : the other one helix):
1) Original DNA
A-t
T-a
G-c
A-t
A-t
T-a
T-a
C-g
2) Separation
A- -------------------- -t
T- -------------------- -a
G- ------------------- -c
A- -------------------- -t
A- -------------------- -t
T- -------------------- -a
T- -------------------- -a
C- -------------------- -g
3) Duplication by DNA-polymerase
A-t -------------------- A-t
T-a -------------------- T-a
G-c ------------------- G-c
A- t -------------------- A-t
A-t -------------------- A-t
T-a -------------------- T-a
T-a -------------------- T-a
C-g -------------------- C-g
So, when one knows what is on an helix, one straightforward knows what is on the sister helix of the DNA molecule.
With the value « 1 » given to G (or C) and the value « 0 » to A (or T), the a.m. digital information « 01100110 » would be encoded by the oligonucleotide « AGGAAGGA » and the coupled oligonucleotide « tccttcct » would be added using PCR to form a DNA strand.
Wouldn’t it be simpler : just play with 2 couples of bases and get directly 2 copies of the information at first encoding ?
If the present idea of digital data encoding into DNA is just to encode one DNA strand (1/2 DNA), the question remains if we wouldn’t lose in storage capacity (storage in 1/2 DNA but without the compactness reached by the chemical bounds in the double-helix) ?
Bob, good question. Considering that at this point there are two strands joined by combinations of 6 molecules in order to encode data, one wonders. Digital data used to be considered 1s and 0s, but there is no reason that it can't be hexadecimal, though I have no idea what each of the 6 would represent. However, there would probably be redundancy built into the storage, thus if one molecule was unreadable then the other could be read.
And I say "at this point" because it was announced last week that scientists have found more data embedded in DNA that they never suspected. I do not know what constitutes this new additional data encoding. Would be interesting if it was genetic memory.
I'm not a specialist in the field of your question but think that the link below and references given there would be useful.
https://en.wikipedia.org/wiki/DNA_digital_data_storage
Very interesting question. There are good resources on the subject including a recent article in NATURE.
How do you encode digital data — which is made up of 1s and 0s — in the building blocks of DNA?
LC: Interestingly, DNA already has a digital “flavor,” as it has four bases and molecules that “stick” to each other in a very programmable way. So the first step in storing digital data into DNA is to map strings of 1s and 0s into strings of As, Cs, Gs and Ts. Next, the DNA sequences are actually “manufactured” chemically, in a very parallel way. Our collaborator Twist Bioscience has a silicon-based DNA synthesis substrate that can make many different sequences in parallel. After the DNA molecules are manufactured, they are put in a test tube and dehydrated. And if protected from light and heat, they can last a long — and I mean very long — time.
http://www.imperialvalleynews.com/index.php/news/latest-news/9987-researchers-break-record-for-dna-data-storage.html
Please see the links for more info.
Goldman, Nick et al. “Toward Practical High-Capacity Low-Maintenance Storage of Digital Information in Synthesised DNA.” Nature 494.7435 (2013): 77–80. PMC. Web. 12 July 2016.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3672958/
http://www.extremetech.com/extreme/134672-harvard-cracks-dna-storage-crams-700-terabytes-of-data-into-a-single-gram
http://www.storagecraft.com/blog/history-of-data-storage-and-backup-part-1-the-writing-on-the-wall/
Dear Colleagues,
Good Day,
"New record for storing digital data in DNA
The DNA in every cell of your body houses an unfathomable amount of information. Harnessing such storage capabilities for the next generation of digital data storage has been the subject of studies for years, and now a team made up of researchers from Microsoft and the University of Washington has broken a new record, managing to store and retrieve 200 MB of data on strands of DNA.
We're getting better at shrinking the physical size of data storage devices while simultaneously increasing the stoarge capacity, with hundreds of gigabytes of data squeezing onto devices that fit in the palm of a hand. But far more data is produced each year than our current technology will be able to keep up with as the world's total data heads towards an estimated 44 trillion GB by 2020.
Unfortunately, even the best of our current range of devices are only relatively short-term solutions to the problem. Hard drives, and optical storage such as DVDs and Blu-Ray discs, are vulnerable to damage and degradation, with a life expectancy of a few decades at best.
Scientists are increasingly looking to nature's hard drive, DNA, as a potential solution to both the capacity and longevity problems. As our own bodies demonstrate, DNA is an incredibly dense storage medium, potentially squeezing in a mind-boggling 5.5 petabits (125,000 GB) of information per cubic millimeter. By that measure, according to University of Washington professor, Luis Ceze, all 700 exabytes of today's accessible internet would fit into a space the size of a shoebox.
You could then tuck that shoebox away in a vault for thousands of years, and the DNA-stored data would remain intact. As evidenced by fossilized remains of woolly mammoths, which have been found to still contain traces of the animals' genetic code thousands of years after they died out, DNA is incredibly hardy and capable of storing information for millennia under the right conditions.
While we won't be using DNA-based hard drives to store vacation snaps in the near future, this latest project is a leap towards more efficient archival technologies for organizations that deal with huge amounts of data. The Microsoft/University of Washington team were able to store, among other things, the Universal Declaration of Human Rights in over 100 languages, the top 100 books of Project Gutenberg, the Crop Trust's seed database, and a HD music video (OK Go's This Too Shall Pass). The data, 200 MB in total, took up less physical space than the tip of a pencil.
In a world where digital data is commonly measured in gigabytes and terabytes, 200 MB doesn't sound like a whole lot, but previous research has only managed DNA data storage on the scale of kilobytes. In 2012, for instance, Harvard geneticist George Church managed to encode his e-book onto DNA, preserving 700 kB of html text, images and formatting instructions – before making 70 billion copies of it.
The UW and Microsoft team, collaborating with Twist Bioscience, were able to encode the data onto the DNA strands by taking advantage of the similarities between DNA's natural code and the binary language of computer code."....
Please, read the rest of the article for more detail...
http://www.gizmag.com/dna-data-storage-record/44273/
Exactly how is digital data encoded into DNA?
As I understand it and to be clear, there is no question here of encoding digital data into existing DNA, e.g. of a bacteria, cell, …..(as one could think at first quick reading). Neither is it intended to use the elements of the cell machinery like RNA transcription, etc.
The idea here is to produce synthetic DNA (which encodes digital information) using an oligonucleotide synthesis machine. The information can then be retrieved at will using a DNA sequencing machine.
Why DNA and not another synthetic chemical molecule ? Because, besides its major target (storing information for life and reproduction) nature has endowed DNA with 3 very interesting “lateral” features (also of high necessity for a good storage of information for life and reproduction) :
- Capacity to wind up into an extremely compact volume : each one of our cells contains a DNA molecule with 3.109 bp (base pairs).
- Capacity to hold its integrity during a very long period (in cold, dry and dark conditions) : see DNA of extinct species.
- Capacity to be copied many times (nearly) without loss of information : this is vital for living species and inherent to its double-helix structure (so with the PCR – polymerase chain reaction - method, it is possible to multiply DNA strands ad libitum).
However, using DNA to store digital data has a drawback : retrieving the information with a sequencing machine must be performed base after base. It takes time. But technology development will most probably reduce this drawback. And it is also not bad to ascertain an high degree of reliability of the retrieved data. To be sure, several (4) copies of the strands can also be systematically made at encoding stage. This would allow error free storage for millions of years.
So synthetic DNA is used for its outstanding storage and reliability features.
As there is no question of storing the information into bacteria, cells, …., devices like graphene nanopore devices are developed to allow the storage of the produced synthetic DNA, as is done with books in a library. But as mentioned in a post above, the manufactured DNA-molecules can also be put in a test tube and dehydrated.
How the data are encoded ? As already mentioned in other posts, a correspondence is made between the Boolean « 0 » and « 1 » and the 4 bases A, T, C, G (= Adenine, Thymine, Cytosine and Guanine) of DNA.
For instance : T and G are given the value « 1 » and A and C the value « 0 ». So the digital information « 01100110 » would be encoded by letting the oligonucleotide synthesis machine synthesize the oligonucleotide « ATGAATTC ». This oligonucleotide can then be read by a DNA sequencing machine to restore the initial information « 01100110 ». Note that there also exists a possibility of base 3 (« 0 », « 1 », « 2 ») encoding.
What I don’t understand is why to give arbitrarily the value « 1 » to two bases and the value « 0 » to the two others no matter how ? Of course, once the choice made, the information will be univocally stored.
However, one could do it simpler : just give the value « 1 » to G (or C) and the value « 0 » to A (or T) or vice-versa.
Indeed, one wants to store DNA which is a double-stranded helix to use its compact storage capacity. So, why not use a basic characteristic of DNA : the fact that the 4 bases are coupled 2 by 2. G is (chemically) coupled with C and A with T. This is the essence of creating a copy of DNA at cell division. After the two strands are being separated, DNA polymerase adds the complementary bases to the existing bases of the original DNA. For instance (capitals : one helix ; lowercases : the other one helix):
1) Original DNA
A-t
T-a
G-c
A-t
A-t
T-a
T-a
C-g
2) Separation
A- -------------------- -t
T- -------------------- -a
G- ------------------- -c
A- -------------------- -t
A- -------------------- -t
T- -------------------- -a
T- -------------------- -a
C- -------------------- -g
3) Duplication by DNA-polymerase
A-t -------------------- A-t
T-a -------------------- T-a
G-c ------------------- G-c
A- t -------------------- A-t
A-t -------------------- A-t
T-a -------------------- T-a
T-a -------------------- T-a
C-g -------------------- C-g
So, when one knows what is on an helix, one straightforward knows what is on the sister helix of the DNA molecule.
With the value « 1 » given to G (or C) and the value « 0 » to A (or T), the a.m. digital information « 01100110 » would be encoded by the oligonucleotide « AGGAAGGA » and the coupled oligonucleotide « tccttcct » would be added using PCR to form a DNA strand.
Wouldn’t it be simpler : just play with 2 couples of bases and get directly 2 copies of the information at first encoding ?
If the present idea of digital data encoding into DNA is just to encode one DNA strand (1/2 DNA), the question remains if we wouldn’t lose in storage capacity (storage in 1/2 DNA but without the compactness reached by the chemical bounds in the double-helix) ?
Dear Mr Skiles .
I have very little idea about DNA(funny !!!!) , digital encoding should be easy since the ATGC are fixed bases . They can be represented in discrete states may be all 4 in one by using phase attribute to the values/bases.
As against 2 states (0/1) , 4 can be better in , radix changes to 4^n and you can larger volumes of data for storage.Eg: music , the lower frequency bands(bass
My concern is the "long term storage". DNA is very fragile and while it can last a long time, usually it is damaged and has to be pieced together. Heat and other factors mean that anything stored to DNA would have to be kept in a controlled environment. It seems like the recent laser-etched crystal quartz glass media, said to have an estimated lifespan of 100-300my and resistance to heat up to 3500 degrees for 2 hours would be more reasonable and reliable method of storage (https://www.engadget.com/2012/09/26/hitachi-announces-permanent-quartz-storage/).
What about the machines to accurately and quickly read data encoded to DNA?
C. C. Ray, New York Times, 6.3.2008, “But even with these advanced technologies, DNA appears as a string rather than being resolvable into the individual units from which it is composed.”
Nice response by Guibert. To the question why to give arbitrarily the value « 1 » to two bases and the value « 0 » to the two others no matter how? It is also a nice question, but I guess it has to do with Chargaff's rule that states that DNA of all organisms should have a 1:1 ratio (base pair rule) of pyrimidine and purine. Probably, they wanted to keep that principle at the time that you ensure around 50% of GC content, with ensuring optimal performance on sequencing platforms. Too much GC or AT can be problematic. I think it makes sense while additionally, you convert the DNA coding system into a binary of zeros and ones that computers use. Interesting question and response.
natural DNA consists of helical structure of two strands. each strand is made up of a sequence of bases A , C, G, T these bases could be encoded into binary A=00, C=01, G=10, T=11. on the other side (biology) we can design a biological DNA starting from a series of bits 0001011001101011 then converting to DNA bases using conversion rule after than we obtain a DNA sequence, this sequence could inserted into a real DNA sequnce with the help of predesigned primer wich starts the replication process in a chemical solution containing free nocilutides with help of some enzyme. this is the way we can insert information into real DNA. the resulted piece of DNA could be inserted into living cell which in turn could transfer the information to another generation. DNA may be the solution for storage in the future where 1gram of DNA is able to stor 100,000,000 Tera Byte of information.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Biology
- Biomolecules
- Nucleic Acids
- DNA