What is it though?
NOT to be confused with computational biology. That’s the very first thing that you should know. Computational biology is about creating models of molecules in a “normal” computer. Just like you can use CAD software to 3D print stuff, you can also use other programs to visualize the structures of proteins, for example.
The point is that computational biology doesn’t involve the actual use of biomolecules, and thus, it doesn’t have to do much with what I’m about to tell you.
Knowing this, biological computing happens when we use biological molecules to mimic technology. I consider it to be a subfield of synthetic biology, and I would define it as:
Using biomolecules like DNA or proteins to perform logical operations, or store digital data
Thinking of biology as a set of tools that can be leveraged for other, before unimaginable, purposes, is what I think biological computing is about.
Another perspective that we could stumble upon is thinking of unmodified organisms as entities that are already computers in some way.
But before we get too philosophical, what if I told you that all the world’s data can be stored in a car’s trunk if it’s stored into DNA? or that biological computers could overcome some limitations that quantum computers haven’t? that one tiny biocomputer is 1000x faster than a supercomputer?
In this 101 guide to biocomputing, I’ll go through all of these exciting topics, and more. Each section also contains some challenges that need to be overcome, as well as my future vision for the technology, and why they are useful for humanity in the first place.
Before the BioLaptop
We’ve mentioned that biocomputing is all about emulating technology, so I think it’s essential to first understand the basics of traditional computing. Don’t worry, I’m not that much of a “techy” person, so I won’t be going too deep.
The classical computer
Although by the end of this article, we may all disagree, Google defines the word computer as “An electronic device for storing and processing data, typically in binary form, according to instructions given to it in a variable program”.
This means that smartphones, tablets, smartwatches, and laptops are all examples of computers. The question now is: how do they work?
Well, the so-known chips can be broken down into transistors, logic gates, and basic modules. Using a bottom-up approach, here’s how this works:
Transistors: tiny devices that regulate the flow of electrons in a wire and amplify certain electronic signals
Logic gates: they can be made out of transistors. They have more than one input and one output, which are related to a specific logic
Circuits: in here, input values proceed through a sequence of gates, to then have an output
I like to think of these as 3 different levels of organization which are the hardware for computers. The software is binary code (1101101000), where the 0s refer to no flow of electrons, and the 1s are the opposite.
This said, we can go deeper into logic gates, which will be an essential concept for biocomputing. Logic gates follow something called Boolean logic. It’s as simple as yes/no, on/off, true/false. The main Boolean Operators are:
AND: all the inputs need to be positive to have a positive outcome
OR: if either of the inputs is positive, the outcome will also be positive
NOT: it reverts the input. If the input is negative, the outcome will be positive
At the same time, these are also logic gates, and they can combine to create other logic gates. In the context of computers, we work with 0s and 1s, and the symbols are:
Biologists discover and engineers create. Discovery is to exploration what engineering is to creation. It’s said that it’s easier to understand electronics because we created it. However, with biology, we just started studying life after it had a few billion years head start.
In my opinion though, we have come to an era in which we don’t know all the secrets that life holds (yet), but our knowledge is enough to start innovating.
Furthermore, biology can get messy. This is why engineering principles could be such a great tool to think of cells more like computers, than burritos. It will allow us to have more reproducibility and predictability of the outcomes.
The principles that we should take into account for biological engineering are the following:
- “LEGO” principle: the different levels of organization in life are your building blocks. If you want to engineer tissues, cells are those building blocks. Proteins would be the blocks for cells, and so on
- Reproducibility: this is harder in biology, but key in engineering. Robotics and ML can help with this by automating lab procedures and better designing experiments
- Testing: without Key Performance Indicators, your project can go wrong. However, it can be challenging if you don’t know what you’ll discover. Ask “What can be measured?”
- Borrowing from other disciplines: just like borrowing the fundamentals of engineering from other disciplines to harness biology
- Reinventing the process itself: breaking the problem down into parts, and then breaking the process down into steps. Think long-term, to project and to plan forward
As mentioned before, biocomputing is a branch that emerges from synthetic biology. Synbio applies these engineering principles to give logic to cells, turn them more into computer-like beings.
The way we do this is by re-thinking genes into logic gates, and ultimately into genetic circuits. Here’s a bit of how it works:
In biology, we have different “biological parts” that we can use to achieve what we want. Even when the symbols may not look very similar to what we’ve just seen in computer logic, most of them have the same functions. Let’s take at some of the most common and essential parts:
Promoter: DNA sequence that’s next to where a gene. It helps to transcribe a gene
CDS: coding region of a gene
Primer Binding Site: region of a DNA sequence where single-stranded primer bind to start replication
Restriction Site: located on a DNA molecule containing specific sequences of nucleotides, which are recognized by restriction enzymes
Just for now, we don’t need to know what each one of these parts does. What is important is that we understand how their logic resembles the digital one that we’ve previously seen with logic gates.
For example, a NOT gate in biocomputing would look something like this:
1. A promoter is turned on because of an environmental condition
2. This promoter then turns on an inhibitor
3. That inhibitor then deactivates another promoter, which would have activated a gene otherwise
In other words, when the input is positive (first promoter is activated), then there will be no output (the last gene won’t be active)
A NOR gate would be a very similar one, only that if any of the two first promoters is activated, then the output will be 0
Having been able to create logic gates by using DNA means that we have the building blocks to achieve much more complex tasks. Indeed, scientists have been able to solve interesting problems using biomolecules, including the Salesman Problem, and have even done Machine Learning with DNA! 🤯
I’ll call Leonard Adleman the father of biocomputing. He was the first to demonstrate the use of DNA to solve the seven-point Hamiltonian path problem in 1994.
The computation in this experiment was done at a rate of 100 Teraflops or 100 trillion floating point operations per second. To have a reference, the world’s fastest supercomputer, Earth Simulator, runs at just 35.8 Teraflops.
The way that a biological computer actually works is by first requiring an input and output signal. The input normally refers to environmental conditions, while the output is the presence or absence of a certain molecule.
In the case of the Salesman problem, the sequence of the DNA actually determines the answer.
Not so fast, really. DNA might have been faster than a traditional computer when finding the answers to the problem. However, the major challenge for biocomputing is what happens before and after.
Indeed, the problem for Adleman was to extract this unique solution. He had to amplify the DNA sequences using PCR, then sort the set of strands by length using gel electrophoresis, next he did affinity purification, and finally, the strands Adleman was left with were then sequenced to reveal the solution.
Interestingly, extracting the solutions to the problem isn’t the only problem. It has been estimated that if this problem was scaled up to 200 cities (instead of 7), then the weight of DNA required to represent all the possible solutions would exceed the weight of the earth!
Another challenge to overcome is the very likely-to-occur errors when sequencing or synthesizing DNA.
This is why some experts don’t have the vision of replacing traditional computers with biological ones. Instead, the most “logical” application would be for biology itself.
To create molecular computing machines that can analyze situations in cells, and then synthesize molecules to deal with them, or use the molecular self-assembly of DNA to build complex molecular structures, which could impact nanotechnology.
Not the most interesting, personally. However, the parallel interactions of peptide sequences and antibodies have been used to solve some problems.
This model of computation has also been shown to be computationally universal (or Turing complete).
An advantage over DNA computing is that the peptide-antibody interactions are more flexible with recognition and affinity.
The main limitation is the availability of specific monoclonal antibodies required by the model.
This specific topic doesn’t have to do a lot with what I consider to be biocomputing. I just think it’s important to mention it because it’s still really mind-blowing.
Wetware is basically this idea of using cells such as neurons to perform computations. Yes, your brain is a computer, but can it be a computer? I mean, could we start using brains as actual external computers?
Unlike conventional materials which operate in binary, a neuron can shift between thousands of states, constantly altering its chemical conformation, and redirecting electrical pulses through over 200,000 channels in any of its many synaptic connections.
This may bring a lot of advantages depending on the applications that we want to give it.
Again, the focus of this article won’t be on wetware, and I think that’s exactly the reason why needed to touch on the topic: to be able to differentiate between the two.
Artificial Intelligence is already a tremendously hot but complex topic. What if we were *already* able to do this with biology?
Researchers from Caltech announced that they have developed a neural network made from synthetic DNA. The network “learned” how to correctly identify handwritten numbers, a very common task when one first learns how to do Machine Learning.
The way this worked in the molecular world was by looking for certain concentrations of molecules and producing specific reactions when it found them.
Potential applications for this include “teaching” our microbiome to speak up and tell us how we can eat better to help it do its job, or just working with other types of cells in a much more efficient way.
The main challenge that biological computing is facing right now is the cost of reading and writing DNA, as well as the lack of preciseness that these technologies still have.
Another point that I always keep in mind is the clear complexity of working with biomolecules, then passing that information on to traditional computers, and vice-versa. Is there possibly a way in which we can skip DNA sequencing, or make it more seamless at least?
Like the experts have mentioned, I think that biological computers won’t be replacing the ones that we already know. Biology will keep on working for biology, but that only means that will be creating more effective treatments and more accurate diagnostics, by programming life in a better way.
One of the questions that I still have is: will synthetic biology be the ultimate solution to many health problems, or will nanotechnology actually come to replace it in a near future?
DNA Data Storage
For sure, biocomputing sounds exciting. Nevertheless, what I think is going to be more feasible, at least in the short term, and when trying to compete against electronics, is DNA data storage.
As a start, we should know that DNA already stores data, a loooot of data. So much that it contains all the necessary information to create a human being. Part of engineering biology is harnessing this capability to store digital data. Scientists have already been able to store all Wikipedia’s information into DNA, as well as movies, books, and bitcoins!
There are a ton of advantages when storing anything in this way. One of them, is that these chemicals can last for up to hundreds of thousands of years, in the right conditions. This is the reason why researchers have been able to study the genetic material of ancient humans, or mammoths.
Another great advantage is that DNA is the densest known storage medium in the universe, (just based on the laws of physics). In theory, we could store every bit of datum ever recorded by humans in a container about the size and weight of a couple of pickup trucks.
Wait… so every cell in our body also contains DNA. Does that mean that we could be having “thumb drives” soon?
Great question! (which I hope you’ve actually asked). So, as of now, and as far as I’m concerned, no one’s actually done this. However, there are other cool objects apart from thumbs that people have used to store data, using DNA.
These include plants, and 3D printed stuff!
Data storage crisis
“The quantity of digital data is doubling approximately every two years. The ability to store this data is not keeping pace with its growth. There is a drastic need for a new storage medium that efficiently and accurately stores data.”
Is it possible that DNA fights this crisis?
The science behind
Conceptually, storing data into DNA is simple. There are 6 main steps that one should follow:
Translate binary data into the DNA language. This means going from 1s and 0s to As, Cs, Gs, and Ts. The equivalence would look something like this:
00 = A
01 = C
10 = G
11 = T
I’m totally making this up, there may be companies who work with different equivalences, like having A = 11, but the principles are the same.
Some researchers have found that working with only 00s and 11s is easier and less prone to errors.
Now that we know the DNA sequence, we can print it, which means synthesizing it: having it physically.
This can be done in many different ways, but that is the topic of another article. Also, this is the step in which things could go wrong (there could be “printing” errors)
The DNA is stored for later usage
When it’s needed, DNA is retrieved
The DNA is read. This means that we’re now translating the physical DNA letters into a digital sequence, that is still written in the DNA language
The sequenced DNA is converted back into binary and is readable by a classical computer
DNA of Things (DoT)
YAAAS! This is one of the most exciting ways to do DNA data storage, in my opinion. You may recognize the name because of its electric equivalent: Internet of Things (IoT).
Well, before you get disappointed, DoT is not very likely to help you turn on your refrigerator any time soon. Still, that doesn’t mean it isn’t shockingly great.
DoT actually refers to storing data into DNA, and then inserting that DNA into an object. The best-known example of this, is the 3D-printed bunny that contained the instructions to 3D print that same bunny.
If that makes any sense, what we would be doing, is the following:
- Know the binary sequence that codes for the digital 3D design of a bunny
- Translate that into a DNA sequence (As, Cs, Gs, Ts)
- Synthesize that DNA (print it)
- 3D print a bunny (or any other object)
- Encapsulate DNA into the bunny
- Take a sample of the bunny (it can be any tiny piece of plastic that forms the bunny)
- Extract the DNA from that tiny piece
- Sequence that DNA (read it)
- Translate that DNA sequence into a binary sequence using the corresponding equivalence that we used at the beginning
- Have a computer read that binary sequence and do whatever we need to do with it ;)
That’s what the DNA of Things is about. Thinking of this in a little bit of a romantic way, isn’t it awesome to think that if the information that we store into 3D-printed objects contains the instructions to 3D print those same objects, we would essentially be emulating biological reproduction?
If you were to store something important into DNA, and save it in a 3D printed object, what object would that be? Tweet your answer and tag me! (@AnaSofiash)
Plants & other organisms
The first practical idea about using plants as storage media was proposed by Karin Fister and Iztok Fister Jr. in 2013 when the two were still undergraduate students!
They first encoded a “Hello World” computer program into a DNA code, synthesized it and cloned this Coded DNA into a plasmid to be used further introduced into Nicotiana benthamiana plants
This was all done without affecting the plants’ vigor and fertility, which is an important factor to consider for sure.
Another question that may arise is: how long will the information last, given that plants die sooner or later? Well, their approach demonstrated that artificially encoded data can be stored and multiplied within plants. That is a win, but perhaps, if we wanted to store important stuff, we’d have to look into the plant’s lifespan as well.
The great great problem that I see for DNA Data Storage, is the fact that it can’t be used in our day-to-day lives, at least not yet. The reason for this, is that I’m sure that I’m not the only person on the planet who doesn’t have a DNA sequencer and synthesizer at home, and even if I did, I’d totally prefer using the cloud instead.
What I’m trying to say with this, is that this kind of storage is only kind of feasible for long-term storage as of now. There are just so many different steps which are all so prone to errors, that it isn’t worth it to use it unless you’re saving something important.
In that case, DNA could be a great option. It lasts for long, seriously long, periods of time, it barely occupies space, and it is safe enough.
Moreover, it’s not only the time, but the cost of reading and writing DNA that makes DNA data storage not able to compete against traditional storage methods today.
According to Microsoft, the cost of DNA storage needs to fall by a factor of 10,000 before it becomes widely adopted. While many experts say that’s unlikely, Microsoft believes such advances could occur if the computer industry demands them.
Exciting! That’s all I can say. As I’m getting more into the field, I’m feeling more convinced that DNA Data storage is going to be something that we’ll get to see more and more, as sequencing and synthesizing technologies improve.
I imagine a future in which we literally have family trees in our backyard, which contain photos of our grandparents, or we store secret information into 3D-printed objects (and nobody notices lol).
Living Medicines Could Disrupt Pharma
The science behind the gut microbiome and its intersection with synthetic biology
I’ve previously mentioned — according to what I’ve read from other experts — that most of the applications of biocomputing are and will likely be focused on medicine.
Inherently, we can think of cells as tiny living computers… but it depends on who you ask. There are some aspects of biology that could be considered standardizable, and therefore, predictable. Others aren’t understood enough.
Well, by applying the engineering principles to biology, we should get rid of those problems and start standardizing life. This means that we could be programming cells such that they bring the health benefits that we want and we need, when we want them and when we need them.
I’m talking about microscopic doctors, also known as living medicines, or smart medicines. This term comes from the idea that the conventional medicines that we know of are chemical substances themselves. Our bodies react to them, but they don’t react to our bodies.
Living medicines are those cells that are genetically modified in order to achieve a specific function. The main characteristic that defines them is the ability to react to their environment, and produce a certain substance in return.
Remember the logic gates and electrical circuits that we were talking about and how we can also create genetic circuits?
Well, living medicines essentially follow that logic and instructions.
To exemplify this, I’d love to include some examples of High School iGEM projects, that have achieved fairly outstanding things.
My favorite one, is Tea-HEE, an engineered ecoli bacteria that can be used as a probiotic. It will respond to PCA, a tea metabolite derived from tea intake, and then produce 5-HTP, the precursor of serotonin. The problem that these guys aim to help to solve is depression.
Another great example of living medicine is SEHS-China’s project. “A Histamine Sensor in bacteria or cell-free systems, which can emit a fragrance when histamine level rises on the skin and nasal mucosa and also release rDAO, a histamine oxidase, to alleviate symptoms”.
In order to deal with the hay fever, we first considered to find the pollen causing the allergy and use engineered…
What does quantum computing have to do with biological computing? Definitely a question that I never thought I’d have to answer.
The truth is, that both of them have shown enough potential to exceed the power of conventional digital computers, provided that technical difficulties are overcome.
Just as with other topics in this article, quantum computing deserves its own guide and explanation. The most basic concept that I am able to explain right now is that quantum computing uses characteristics of subatomic particles, like superposition and entanglement to compute not only with 1s and 0s, but with states in between those two, allowing us to solve complex problems, rapidly.
Some have argued that biocomputing could beat quantum because it doesn’t need a super big computer, and it doesn’t require that computer to be at such some cold temperatures, like quantum computers do.
On the other hand, some people go one step further, and say that it’s the combination of DNA and quantum computers that can really help us thrive.
I’ll definitely do more research on this specific question, and write an article about it, once I’ve gotten a deeper understanding of both technologies. Stay tuned for that!
Shoutouts to companies
If you’ve read other of my articles about biotech, you know that I’m rather enthusiastic about taking innovations out to the market. This section is dedicated to those innovators in super-big or small companies that are making a difference by accelerating the development of biocomputing technologies.
Grow Your Own Cloud
Having proved that plant data storage is possible, GYOC is a startup that does exactly that: it stores digital data into DNA, and then puts it inside of plants. There is no other thing in the universe that makes more sense to me at this moment.
The data crisis tells us that we are probably creating more data than the one we can store. The ecological crisis tells us that we need more plants producing oxygen. What solves these two problems at once? Storing data into plants!
A simply beautiful solution!
Microsoft’s Station B
It’s not necessary to place this at the top, since it clearly stands out on its own. You also know that something could be “The Next Big Thing” when a unicorn is working on it.
“The Station B project builds on over a decade of research at Microsoft on understanding and programming information processing in biological systems, in collaboration with several leading universities. The name Station B is directly inspired by Station Q, which launched Microsoft’s efforts in quantum computing, but focuses instead on biological computing.”
Synthetic biology’s dogma is: Design, Build, Test, Learn (DBTL). Microsoft does this by designing biological programming languages, compilers that translate high-level programs to DNA code, testing biological experiments using lab robots, and learning by analyzing their experimental data with AI.
If I were to mention one of the places where the future of biotech is taking place, Microsoft’s Station B would definitely be on the list.
Last, but definitely not least, there is Twist. This company is already an outlier in all-things DNA. Of course that it had to work on DNA data storage as well. Fun fact, Station B is partnering with them!
Twist has made the DNA synthesis process faster, cheaper, and at a higher throughput than most other synthesis companies. The role that it has in biocomputing is by improving one of the elementary technologies that are needed for it to occur in the first place.
Actually, Twist has very recently partnered with Illumina (a company focusing on DNA sequencing) as well as Western Digital (a data storage company) to help Microsoft advance DNA data storage.
This is truly exciting news. It’s like the perfect alliance.
In a Nutshell
That was a long article, but probably a short Introduction to Biocomputing. This is part of my “Learn” phase in my biocomputing journey, which means that I’ll be releasing more interesting and deeper content very soon.
As of now, what I can say is that I’m equally excited about the medical applications and the technological ones. It’s said that — just as with quantum computers — we won’t be having a biological computer any time soon, since these will actually be used to solve very specific and complex problems.
Nevertheless, I wouldn’t mind giving DNA data storage a try as something that substitutes hard drives, for example.
I’m also tremendously curious to know how we can overcome the problem of having so many intermediate steps between the storing and retrieval of data.
Furthermore, I consider that smart solutions to the data storage problem, like the one that GYOC is developing, are ones that we will want to be implementing without a doubt.
In the end,
the best way to predict the future is to create it — Abraham Lincoln
Hey! I’m Sofi, a 16-year-old girl who’s extremely passionate about biotech, human longevity, and innovation itself 🦄. I’m learning a lot about exponential technologies to start a company that impacts the world positively 🚀. I love writing articles about scientific innovations to show you the amazing future that awaits us!
Twitter | LinkedIn | Website | Podcast | Newsletter