Science Friday - Using DNA To Boost Digital Data Storage And Processing

Episode Date: September 9, 2024

You might be familiar with a gigabyte, one of the most popular units of measure for computer storage. A two-hour movie is 3 gigabytes on average, while your phone can probably store 256 gigabytes.But did you know that your body also stores information in its own way?We see this in DNA, which has the instructions needed for an organism to develop, survive, and reproduce. In computing storage terms, each cell of our body contains about 1.5 gigabytes worth of data. And with about 30 trillion cells in our bodies, we could theoretically store about 45 trillion gigabytes—also known as 45 zettabytes—which is equivalent to about one fourth of all the data in the world today.Recently, a group of researchers was able to develop a technology that allows computer storage and processing using DNA’s ability to store information by turning genetic code into binary code. This technology could have a major impact on the way we do computing and digital storage.To explain more about this technology, SciFri guest host Sophie Bushwick is joined by two professors from North Carolina State University’s Department of Chemical and Biomolecular Engineering, Dr. Albert Keung and Dr. Orlin Velev.Transcripts for each segment will be available after the show airs on sciencefriday.com. Subscribe to this podcast. Plus, to stay updated on all things science, sign up for Science Friday's newsletters.

You might be familiar with hard drives as a way to store your data. But did you know that DNA can store digital information in its own way? DNA has string of letters, ACT, and G. So you just go letter by letter and convert that into these digits. It's Monday, September 9th, and this is Science Friday. I'm SciFriar Radio Fellow, Valeria Diaz. Each one of our cells contains the equivalent of about a gigabyte of DNA data. And with an estimated 30 trillion cells in our body,

all that DNA adds up to roughly 30 trillion gigabytes of storage. That's enough to encode about one-fifth of all the data in the world today. In recent years, researchers have developed technologies that tap into DNA storage capabilities. And now, researchers are going beyond storage and using DNA as the basis for computers. Here's SciFRI guest host Sophie Bushwick with more. I'm joined by two professors from North Carolina State University's Department of Chemical and Biomolecular Engineering. engineering. Dr. Albert Kjong and Dr. Orlin Vellev. Welcome to Science Friday. Thank you so much for being here. Thank you, Sophie, for having us. It is a pleasure to be on this very interesting discussion.

Thanks. And how does DNA store information? The simplest way to think about it would be DNA is a string of letters, A, C, T, and G. And so you can have any length string of letters that you want, and you can have many of these of these strings. The simplest way to convert the letters into binary or zeros and ones would be an A could be a zero zero, a T could be a zero one, a G could be a one zero, and a C could be a one-one. And so you just go letter by letter and convert that into these digits. And what about if you want to go beyond storage and use DNA to process information as well? How does that work? So this has actually been relatively active field for over 20 years since Leonard Ampley.

Bowman first created the first computation with DNA. I mean, there's many different flavors of this computation. You could use enzymes that can recognize and chew up certain pieces of DNA that have certain sequences. There's types of computations that use interactions between different DNA molecules to bind or unbind each other and execute logical operations that way. So the past two decades have actually generated many critical. creative versions of computation.

What about your work? What did your latest study focus on? There's been two decades of work in DNA computation. And in parallel, but somewhat disjoint, there's been also work on storing information in DNA. What we wanted to do was create something that was compatible with both storing and computation, basically try to create a early full computer, something that we think could help spark. the imagination of young scientists out there that might be thinking about getting into research and engineering and science. And so our focus was really on, can we create something that can both store, but it's also warm enough, kind of flexible enough, to be used dynamically

for things like computation. Can you give an example of a computation that it could be used for? Yeah. So one of the computations I found really fun was work over a decade ago from Princeton, they computed a chess problem. And so this was one type of computation that we emulated, and then another was Sudoku. These puzzles are basically, have kind of similar rules. So basically, it's asking where can you put different chess pieces on a chess board so that they don't attack each other or don't attack a certain piece? Or in Sudoku, where can you put zero ones, two, threes, so that only one digit shows up

in a row column at any one time or every row, every column adds up to six, things like that. So you have certain puzzles, certain board configurations that you're searching for. And if we want to use DNA as a computer, it has to be able, like you said, to store this information and also process it. But some of the techniques you talked about for processing information, like using enzymes to chew up the DNA, that doesn't seem to be possible if you want to also store the DNA. So how did you get around that problem? Exactly. I think that's kind of the disconnect that we were trying to find a solution for. So we needed a way to preserve and anchor the DNA without giving up its high density,

information density, but also make it so that you can access the data and compute upon it without destroying the database. So this is where we linked up with Orlin Vellov's group, who pioneered a nanomaterial that maybe he can tell you about. The key discovery was that DNA adhered to this material stably, but allowed enzymes to come in, make copies of the DNA into RNA, and then we could use that RNA to do computations without disturbing the original DNA. Got it. So yes, I'd love to hear more about this nanomaterial. It can be a really pleasure to participate in this project because this is a really multidisciplinary investigation. Some time ago, we got together with my colleague Albert and we were discussing

how we can basically use the innovative tannum materials that we make in my group and we study in order to manipulate and process DNA. And we can just come across this new material, which we call it softendritic solids. It is a fibular material which is made out of biopolymer and it is branched. It has this hierarchical structure. So you have a thicker branch in the middle and then thinner and thinner branches, which come to be nanofibers all around.

And nanofibers tend to be very sticky in physical perspectives. The reason gecko legs can run on any surfaces, that is gecko lizards can run on any surfaces with their legs, is that they have this sticky mats of nanofibers.

So it turned out that now our nanofibers in the new materials, that we're making can be very sticky to DNA. Basically we have a particle of fibroar nature that is combined DNA and in this way we can immobilize

the molecule, we can protect it in physical and mechanical sense and we can even manipulate the whole cluster of DNA that has been collected by using magnetic particles which are also included in the structure. So basically

while Albert has been providing the software in a sense. We have been trying to provide the hardware that is going to allow the whole thing to be protected and manipulated. When you describe the polymers as sort of like the tree branch with thinner and thinner pieces coming off it, it makes me picture the DNA sort of tangled up in a forest of trees, but on a very, very tiny scale. Is that an accurate way to think about it? Well, that is an interesting. That is an interesting. That is an interesting. interesting analogy. Well, if you think of a DNA as, let's say, delicate biological objects, such as a bird,

a tree is an ideal way to protect a bird in the sense that it can fly in. The branch is protected, but then it can still go out. So basically we have access to the inside, but we have protection from the inside when the molecule is hosted within this hierarchical structure. And the other thing that's important about hierarchical. structures of this type is kind of a little bit more scientifically said. They have very high surface to volume ratio. So we use a small amount of material, but we create lots of surface area that is then available

for DNA molecules to bind. So we do not use too much material, but we can bind lots of DNA on those particles. And as I mentioned, we can also add magnetic nanoparticles during the formation. so the whole cluster at the end is going to be magnetic. Got it. I mean, and does that mean that I could plug a computer monitor into the DNA computer and it would theoretically run? No.

Well, actually, yes. You would need an electronic interface. And the timescales of the operations would be very slow compared to what you're used to. How slow? On the order of probably a few hours to enter in a command and then get the result. Okay. So what does this whole setup look like? If I've got my computer monitor, I'm waiting on the results, but what does the DNA computer part of this setup look like? You probably will always need a electronic computer as an interface. And what it would do is basically act as an intermediary

between a very high density DNA setup and database. It would basically process whatever data that you want from it and then display it so that the human could see it. The setup would look something like microfluidics. So you have either tiny, very thin, kind of capillary-like tubing, or it could be microfluidics. That's what we used in this work, but you could also create microfluidic devices that are like little chips that have very small, narrow channels inside. And you could put different databases within these channels and flow basically different solutions containing your enzymes or just water through these channels in order to make copies or execute computations, access the data that you want. You would then flow, say, the RNA copies of the data that you want out of the

nanomaterial that's linked to the DNA. The RNA would come out of that and flow into what we call a sequencer. There are several different technologies. The one that we use is called it Oxford Nanopore. So those RNA molecules would then flow through the nanopore and give off different electrical signals as it passes through that pore and those signals would correspond to the different letters. And you would get that read out that would get sent to your computer. And I know that DNA is a very compact way of storing information. I mean, each one of our cells has about two meters of it and then it's compressed into just about six microns. So how much smaller could we make our computers if they use DNA? for some of this data storage and processing?

Yeah, so theoretically, if you do what's called freeze drying of the DNA, meaning you basically evaporate all of the water away and you're left only with the DNA, it can be very, very compact. You could literally store all of the world's information square foot. And if I can go back to the material aspects of this work, DNA can store lots of information, But it is also a delicate molecule, and it is easy to encode material, to incal information in DNA,

but not that easy to then find the right molecule and pull it out of the rest. So that's why you also need the materials component, which is how do we protect, immobilize, move around, sort out the DNA. So that's what makes this research hopefully interesting, is that it really has all those informatics aspects and molecular and materials and electronics even. That's a really good point because we think about data storage. I mean, I know that it doesn't last forever. A USB drive might only work reliably for, you know, a decade or less. So as a computer and as a storage method, how does DNA compare to other forms of data storage? How long can it preserve itself? One of the really main drivers of the DNA storage field has been not only the incredible information density, but the longevity.

There's been fossils that have been discovered that are a million years old and people have been able to extract DNA from it. I think that there's caveats to that in that the DNA is degraded and you aren't able to access all of the DNA in a pristine condition. However, very simple storage methods can preserve the DNA for a million years. This is one of the key advantages of molecular storage. Theoretically store DNA for thousands of millions of years at near room temperature or in like a household freezer without having to expend very much energy to do that. In comparison, a lot of like media, like you mentioned, a USB stick, or a lot of the long-term storage media like tape,

magnetic tape, these are actually not very stable. I think we often think about inorganic materials as very robust and hardy, right? But they actually have a lot of defects actually just coming out of the manufacturing plants, and a lot of these defects are engineered around in your devices. And over a few years, radiation that naturally comes from space can degrade your devices, just wear and tear from heat, oxidation. And so even the tape storage that's used for long-term archival storage of data, those often you need to copy the material every five to ten years onto a new tape reel. And you have to repeat that every five to ten years.

You've said that DNA could revolutionize computing. What do you see this tech being used for? So I do not really see it as, you know, replacing your laptops or personal computing. But I think that there's a lot of things that are important for our everyday economy and society that rely on computing in the background that we don't know about. We don't see. So things are happening at data centers, these really large buildings with massive energy, land footprints that are executing computations for us. So even when we do like a Google search

flight, where is that computation happening? It's not actually happening on your computer. It's off somewhere else. And there's a lot of these large scale computations that are really important for industry, for academic research. And I think DNA could be very powerful for those types of processes where you need to make very, very complicated and demanding calculations. populations that require a lot of storage but also parallelized computation. Yes, if I may add a little bit of different angle away from informatics, the ability to store and manipulate and deliver DNA and RNA can also find applications in other areas, such as drug delivery, vaccines, plant treatments. So really kind of like combining materials and

informatics in this case, I mean, really have lots of potential future implications, which is still to be understood, probably. And what about the two of you? Where do you see your research heading now? Oh, in so many directions. We actually have a couple projects that are still ongoing, related to just the properties and DNA as a material itself, as well as other types of materials that the VEVILA group has been pioneering and how that interacts with DNA. whether we can protect it for millennia at room temperature, for example. There's a lot of fundamental questions as well that just about how these materials work at the nanoscale that we're also interested in.

I can say that we have been really very inspired by what we have learned from Albert's group in the sense that there are different methods for manipulation of particles on the nanoscale and sorting out in microfluidic devices that was mentioned, but we have been interested also in external fields, electrical fields especially. So this has been really a very productive cooperation in terms of combining ideas from different areas and finding out interesting new applications for both DNA and nanoscience. Thank you so much for joining us. Thank you so much, Sophie, for having us. Thank you. Those were North Carolina State University's Department of Chemical and Biomolecular Engineering Professors,

Dr. Albert Kyeong and Dr. Orlin Vellev. And that's all the time we have for now. A lot of people help make the show happen, including Sandy Roberts, Robin Kasmur, Jordan Smudjik, Charles Burgquist, George Harper, and many more. Next time, we'll talk about the ants that called Times Square their home, and just how many hot dogs a year they can eat. But for now, I'm SciFire Radio Fellow, Valeria Diaz.

Thank you for listening.