Quantum Computing: A Virus’ Greatest Enemy
As made evident amidst the ongoing (and seemingly never-ending) pandemic, viruses spread much faster, farther, and more frequent than ever before. Without losing sight of the tragedy Sars-CoV-2 (COVID-19) has caused, this global pandemic has certainly proven that our ability to detect and treat illness has improved drastically since past disease outbreaks. While it took today’s biomedical researchers less than a year to develop an authorized COVID vaccine, — which was by no means a simply feat — the first licensed Spanish influenza shot did not appear in North America until the 1940’s (twenty-two years after the flu pandemic begun!). Likewise, mapping the outer spike protein structure of the COVID-19 virus took a mere twelve days. This is comparable to the four years it took scientists to perform an analogous structural investigation for HIV. If anything, however, the novel Coronavirus has also taught us that the drug development industry is far from perfect.
Current pharmaceutical advancement processes are very empirical; that is, they rely heavily on observable experiment rather than theory or pure logic. Firstly, biological products/molecules must be manufactured. This might include the antigens, preservatives, stabilizers, surfactants, diluents, and so much more; all vital components to a vaccine. The prime issue lies in an inability to accurately predict the performance of each element beforehand. As each ingredient serves a specific purpose, they must all be tested (both independently and together) in a long, rigorous, and cumbersome process. This not only shatters the cost/benefit ratio, but also leaves the potential complications of a vaccine — even after it begins distribution— up in the air.
I know what you’re thinking.
Why isn’t there a better system?
Before jumping in to the heart of this article, it is important to note that current vaccine manufacturing processes do entail computation, specifically in the design phase. But once these molecules go through subsequent lab testing and clinical trials, researchers are often sent back to the drawing board.
In a perfect world, researchers and scientists could transfer all of the costly and time-inefficient laboratory processes to a complex digital simulation in the earliest stages of drug development. By utilizing pre-existing molecular databases, mathematics could determine which compounds could serve as vaccine components with great precision, accuracy, and efficiency. Put most simply, this largely entails programming a computer to calculate the energy configurations and complex structures of a vast array of molecules. We could then input our factors, screen the database of properties/chemical behaviours, and thus, instantaneously identify a cure.
Sounds great, right? Easy as — but wait! We don’t have the classical computing power to simulate all these complex particles, nor solve the intricate equations that coincide.
Current supercomputers can only exactly simulate around twenty electrons. To fully illustrate this limitation, consider a caffeine molecule with twenty-four atoms. Simulating this molecule with chemical accuracy by conventional means would would require a computer one to ten percent the size of earth. In principle, this is indeed theoretically possible. But it is obviously not feasible.
It seems as though our computational simulations in drug development we would be creating more problems than actually solving them. But here’s the punchline: the (somewhat) novel solution to combatting the novel coronavirus lies in an alternate form of computation. It’s called quantum computing.
Put most simply, quantum computing (QC for short) harnesses the phenomena of quantum mechanics to store data and solve problems.
Let’s back up a little.
Classical computing — the method of computation that allows you to read this article — manipulates zeroes and ones to solve operations. These are otherwise known as bits. Quantum computers, rather, use quantum bits, or qubits for short. Just like classical computers, quantum computers work with zeroes and ones. But catch this, qubits have a third state; a superposed combination of zero and one; that is, until they are measured. This concept is very fittingly known as superposition. It’s important to keep in mind that the quantum phenomena that fuel quantum computation — including superposition — defy all classical logic. Unlike the tangible effects of gravity that we encounter on a daily basis, superposition, entanglement, and interference (don’t worry, I will get to the latter two later) cannot be seen nor felt in the physical world. Such quantum phenomena only arise in the world of subatomic particles. In that realm, the laws of classical physics cease to exist.
Okay, back to superposition!
Take a look at the below animation.
Just like the above skateboarders, quantum particles can be in two places (states) at once. As the eyeball appears, however, only one skateboarder remains. This is also true for quantum objects. Prior to being observed, the quantum bit had relative probabilities of being zero or one. Thus, a single qubit could be described as a linear combination of |0⟩ and |1⟩. When observed/measured, however, it collapses into either of the binary states. Like many complex scientific topics, this is best explained with an example!
If someone was to take a photo of a figure skater spinning around very fast, it would likely appear as blurry photo. But if the figure skater were to be a quantum particle, they would, without fail, either be observed facing directly left or directly right in the photograph. Nothing in between. Once a qubit has been collapsed, it remains in that binary state. So, once the skater’s photo is taken, they will stop spinning entirely, remaining facing either directly left or directly right. As you will see, though, quantum particles can be ‘reset’ back into a superposed state to be functional for operations once again.
Remember the probability phenomenon I mentioned just above? Let’s reimagine the figure skater — a quantum particle, for the sake of my argument — being photographed. Suppose the photographer had programmed their camera to a probability of capturing the skater facing the right 60% of the time, and facing left the remaining 40% of the time. Well, that’s what would occur. The camera’s probabilistic behaviour has been interfered with. Similarly, quantum interference influences the qubits measurement probability, and ultimately, the state it ends up being once collapsed.
Let’s talk entanglement, AKA, “spooky action at a distance”.
Quantum entanglement refers to a single, shared quantum state of a pair of qubits. By changing and/or measuring the state of one of the two members, the state of the other will instantaneously change in a predictable manner. If we were to collapse one of the entangled qubits, we know that the other, no matter the very long distance that may separate them, is measured in the opposite binary state.
Consider two boxes. We know that one includes a red stone, and one includes a blue stone. By opening the first box, we immediately can figure out which rock must be inside the other, as shown below. Although this is a pretty simple (and classical) analogy, quantum entanglement acts in a similar way. It’s all about correlated measurement outcomes.
Despite all all these remarkable phenomena that change the game for all of computation, a major hurtle remains for QC. Quantum computers are very sensitive. Typically, qubits operate at 20 millikelvin, or about -273 degrees Celsius — temperatures that are even colder than outer space. Because of inevitable interactions with the environment — such as changing magnetic and electric fields, radiation, or movement — the ability for qubits to remain in superposition and entangled is often jeopardized. This is known as decoherence, a process that uncontrollably changes quantum particles’ states and causes information stored by the quantum computer to be lost. While decoherence is vital for the quantum measurement phenomenon described above, it’s still a big problem that quantum scientists/researchers are working to combat.
I know that I’m talking up QC a lot, but I want to make it clear that it’s not good for everything. Classical computation is deterministic. It will give you the same, correct answer every time a program is run with the same inputs. Quantum computation, on the other hand, is probabilistic. It will give you an answer, and perhaps, an indication of how correct that answer might be. Run the same program again, you may get a completely different answer. Quantum computing actually only fares better than classical systems when working to solve a very particular category/class of problems. But when it does so, quantum’s capabilities are truly remarkable.
So, I’m sure what you all have been oh-so-eagerly asking yourself, how is all this stuff so revolutionary for pharma and drug development?
First, I want to return to the foundation of quantum computing; the qubit. For two classical bits, they can take the following values, 0 & 0, 0 & 1, 1 & 0, and 1 &1. Two qubits take all those values at once. A pattern emerges; that is, one qubit can take the value of two bits. Two qubits can take the values of four bits, four qubits can take the values of 16 bits, and so on.
Observe the table below.
So, we can conclude that ‘n’ qubits can take the values of 2^n classical information. While 13 qubits is a kilobyte of RAM, 13 classical bits requires a mere byte-and-a-half. To really illustrate this, in 2019, Google’s quantum computer did a calculation in less than four minutes that would take the world’s most powerful supercomputer 10 000 years to execute. This is because of the ability to ‘test’ an exponentially greater number of outcomes simultaneously in the quantum realm. If this sounds truly mind bending, it’s because it is. The extraordinary properties of subatomic particles seem to not make sense when compared to classical mechanics. But the amazing thing is, quantum mechanics actually does a better job of explaining the natural world than any prior theories that stemmed from a classical understanding on the macroscopic scale.
Only 1500 words later, we return to the real purpose of this article — molecular chemistry in drug discovery.
Quantum computation can be leveraged to predict the 3-D structures of protein in target identification, optimize dose and solubility levels for molecules and molecular systems, link appropriate biological information and data points when combatting rare diseases, improve the vaccine supply chain, and so much more.
According to a future casting 2030 vision report, life sciences membership organization, Pistoia Alliance, examined emerging tech that might disrupt the pharma industry’s efficiency, and QC was certainly identified at the top of the list. In January of 2021, the world’s largest private drug company, Boehringer Ingelheim, announced it would partner with Google to use quantum computing in pharmaceutical research and development. This industry giant has also created an internal lab to combine the usage of AI and quantum with their current studies/plans. In that same month, Roche, the world’s largest pharmaceutical company, set for a collaboration with Cambridge Quantum Computing to design algorithms for the infancy stages of drug discovery. “We are scanning the horizon, waiting for the big wave, but we don’t know how big it is going to be, or when it will com,” Mariëlle van de Pol, head of Roche’s quantum task force, said of the technology. “But if you see how much tech companies are investing in this topic, and how quickly the whole landscape is evolving, you realize it will come and it will be a game-changer.”
As pharma companies migrate towards QC, we are seeing a paradigm shift in the world of drug R&D.
While we have come a long way from the mystified mind of Einstein as he attempted to understand quantum physics in what was thought to be a classical world, there is a long, bumpy road that remains in transforming theory into reality. Quantum computing is still in its fragile infancy. There are still a lot of unknowns. One thing is certain, however: quantum computers hold immense disruptive promise, especially in the industry of pharmaceuticals. Although COVID-19 may be soon reaching its end, future widespread disease is inevitable. And as we pivot towards emerging tech to address challenge, quantum computing should play an instrumental role in whatever global health threat might come next.