.svg)
How Quantum Computing Can Be Used for Drug Development
Drug discovery companies can leverage quantum computing today to find compounds to treat and cure diseases. In conjunction with other leading-edge technologies, such as cloud computing, artificial intelligence (AI), and machine learning (ML), they can fully realize the potential of quantum computing for drug discovery by designing new molecular drugs. Quantum computing is significantly faster, much more efficient, and remarkably less expensive than all known alternatives.
As it pertains to drug development, the answer to why is quantum different lies in how very large problems can be addressed as compared to classical computers and even supercomputers. We’ll actually be discussing two different quantum technologies that are both used for computation. Annealing-based quantum computation is in use today and enables drug discovery firms to search chemical spaces a million times larger than would otherwise be the case. Gate-based quantum computing, although not yet ready for commercial use, stands to become even more powerful. Adding just one fundamental information unit, called a quantum bit or qubit, for short, increases the computational workspace exponentially. In other words, adding just one qubit to a gate-based quantum computer doubles the size of the problem we can address. For example, a small 40-qubit device can address a trillion different combinations. And what we’ll basically be discussing here, is searching a very large chemical space of molecules to find the select few that meet some specific criteria. Quantum computers handle large problems naturally.
How can quantum computing help drug development?
Beginning with a specific disease, or perhaps just a specific protein target of that disease, drug discovery companies can already start establishing the profile of a future drug. This drug profile will eventually take into account all the desired properties of this drug, including such considerations as specific protein binding sites, how the drug will enter the body (eg., injection, pill, etc.), how the drug may affect the brain, how much of the drug should be given, what is the target age group, compatibility with other drugs, schedule and duration of administration, packaging and shipping requirements, special marketing considerations, and so forth.
That drug profile is then used to generate a large virtual chemical space, which is when we can then start to see why there is an interest in quantum computing in the pharmaceutical industry. By “large,” we mean that the chemical space includes literally billions of molecules. The vast majority of those billions of molecules, however, are irrelevant to the selected drug profile. Therefore, quantum algorithms such as the Quadratic Unconstrained Binary Optimization (QUBO) algorithm, which is used for combinatorial optimization problems, can search this large chemical space for the specifically selected properties of the drug profile. The results of the QUBO algorithm are relevant to the drug profile in regard to both the desired properties and the desired binding site(s).
By greatly reducing the chemical space, quantum computing significantly reduces the extent of subsequent bench work that needs to be done. Bench work for drug discovery can include tests for toxicity, appropriate dosages, potential costs, and so forth. Completely inappropriate molecules have been removed from the chemical space. Far fewer molecules need to be designed, synthesized, measured, and experimented with. The path to defining and licensing assets is shortened. According to the University of Cincinnati College of Medicine, “a Phase I trial [currently] takes several months to complete” and only “about 70 percent of experimental drugs pass this initial phase of testing.” Decreasing the first statistic and increasing the second statistic presents a clear quantum computational advantage.
How will quantum computing affect medicine?
Quantum computing empowers altruistic goals, such as wanting to treat more diseases in less time than we could possibly do just a few decades ago, and maybe even just a few years ago. Target diseases might currently be of personal interest to researchers, or they might fill in some of the gaps in the present pharmaceutical market. One research goal might be to improve drug efficacy in an area of the market where the existing drugs fall short. Another research goal might be to find smaller molecules with which to improve delivery methods, perhaps allowing patients of all ages to put away their syringes and to start switching to ingesting digestible tablets or capsules instead. Quantum computing allows drug discovery firms to start building their asset portfolios right now with molecules that hold great potential to become life-changing medicines in the near future. Even getting these drugs just into the early stages of their development can shorten the time to put them into production later.
With quantum computing, pharmaceuticals can be brought to market much faster. The actual time it takes to reduce the initial chemical space, from many billions of molecules to mere hundreds of thousands of molecules, takes less than one minute. Of course, there’s much more to a project than a one-minute calculation, though. The quantum algorithm does require preparation -- designing quantum circuits at even small scales is remarkably challenging and time-consuming -- and then there is classical post-processing, as well. Traditional chemistry tools reduce the intermediate chemical space from hundreds of thousands of molecules, perhaps, to mere hundreds of molecules. Instead of drug discovery companies testing many billions of molecules, only the few hundreds of molecules that made it through both quantum and traditional processes need to be synthesized and measured. The project time might be around three months.
It is worth clarifying with a second mention that with the current state of quantum computing, that this is a two-step process. First, a quantum annealer -- a quantum technology that specifically addresses optimization problems -- reduces the initial chemical space from many billions of molecules to hundreds of thousands of molecules. The aforementioned QUBO algorithm runs on this quantum annealer. A distinct second step is needed to reduce this intermediate chemical space down to the final few hundreds of molecules, and this process still uses traditional chemistry tools. Although hundreds of molecules is a broad chemical space, those molecules at least have all the desired properties required by the selected drug profile. This hybrid quantum-traditional approach offers a significant computational advantage over a traditional-only approach, while leaving room for down the road when fully-quantum approaches will become available and the computational advantage will be even more pronounced. Hybrid approaches, besides leveraging quantum computing, usually incorporate other advanced technologies, such as cloud computing, artificial intelligence (AI), and machine learning (ML).
Generally speaking, drug discovery firms are not limiting their searches to molecules with specific properties for specific protein targets, although that is the most important outcome. As alluded to before with a mention of various delivery methods, these companies are also searching for the smallest possible molecules that can fit each drug profile. It’s very difficult, from the outset, to search a chemical space with literally billions of molecules for the desired drug properties. It’s even more difficult, as you may imagine, to find small molecules in such a large space. Like putting bumpers into bowling alley gutters, quantum computing makes it easier for drug discovery companies to find their targets: the smallest possible molecules that have all the desired properties of a selected drug profile.
It’s worth noting that one of the natural advantages of quantum computing is simulating quantum systems. Besides searching a chemical space for molecules that meet specific criteria, quantum computers can simulate the molecules themselves. One potential application of this is repurposing existing therapeutics. One affect on medicine could be skipping new drug development altogether, and repurposing drugs that have already been developed and for which at least some clinical data already exists.
What is the future of drug development?
According to the University of Cincinnati College of Medicine, phase 1, 2, and 3 clinical trials currently take several years, and only about 25-30% of drugs pass all three phases. Each phase tests the efficacy and safety of a drug, but increasing the sample population with each subsequent phase. Phase 1 trials may have less than 100 subjects, phase 2 trials may increase this to several hundred subjects, and phase 3 trials may include thousands of subjects. According to a JAMA paper, the average cost to develop a drug is just under US$\$$1B, with high-end estimates near US$\$$3B. The future of drug development is shorter clinical trials, greater effectiveness, and enhanced safety at a much lower cost.
The method that has been discussed in this blog article thus far uses a quantum technology called quantum annealing. There are actually several quantum annealer providers in the quantum technologies industry, at least three, in fact, that are well-known, and there is also such a thing as a digital annealer. A digital annealer, which could also be called “quantum-inspired,” can solve the same problems as quantum annealing despite not actually being a quantum technology. The primary advantage of both quantum annealing and quantum-inspired annealing is that they are large enough today to solve problems even when chemical spaces include the many billions of molecules that have been referred to in this article. In addition to that, there are presently-known methods to use existing quantum annealing and quantum-inspired annealing devices to work with much larger chemical spaces, if and when needed.
Another quantum technology called gate-based quantum computers, or universal quantum computers, are neither large enough nor accurate enough to be useful today. Unfortunately, they are still in their very early stages of development. However, they are most likely the future of drug development. Like quantum annealing and quantum-inspired annealing, they will be able to reduce large chemical spaces from their initial billions of molecules down to hundreds of thousands of molecules. But, unlike quantum annealing and quantum-inspired annealing, gate-based universal quantum computers will also be able to perform the computational chemistry calculations to further reduce the intermediate chemical space from hundreds of thousands of molecules to only hundreds of molecules. Solving Schrodinger’s equation with gate-based universal quantum computers will provide a computational advantage even over current hybrid quantum-traditional methods.
Reducing the overall time of new drugs to market, first with quantum annealers and digital annealers and later with gate-based universal quantum computers, is critical in this day and age. COVID therapies and vaccines were developed remarkably fast, and yet this historical timeframe remains relatively slow when compared to the quantum methods that we anticipate having in the near future. There is also ongoing concern about these therapies and vaccines, including in regard to their short-term efficacy and their long-term side effects. One of the many promises of quantum computing is discovering fast, safe, and effective medicines for all diseases, including in humanity’s fights against future pandemics. This may be through new drug development or, as noted earlier, through repurposing existing drugs.
Will quantum computing Transform Biopharma R&D?
Disregarding, for a moment, the future capabilities of gate-based universal quantum computers, quantum annealers and quantum-inspired digital annealers are transforming biopharmaceutical research and development in the present day. New drugs can already be discovered faster than they could ever be discovered with traditional methods. Plus, larger problems can be addressed.
Overall, the method of drug discovery is improved. Traditionally, a laboratory may identify thousands of molecules, for example. First note that this example uses a count in the thousands instead of in the billions. Now imagine that the desired properties are the same and the target protein is the same as if using either quantum annealing or quantum-inspired annealing. Yes, the traditional methods work. With time, the chemicals can all be synthesized and measured and tested. New drugs can be found, of course. However, imagine wanting to improve some of the drug’s properties. Or, imagine wanting to add some new properties to the drug. The traditional process becomes a cycle. Drug discovery companies design, synthesize, and measure initially, but desiring improvements of any kind require them to do it all over again.
With quantum annealing and digital annealing, and in the future with gate-based universal quantum computing, this drug development cycle ends. Instead of processing hundreds or even thousands of molecules at a time only to repeat the process for every desired improvement, drug discovery firms can process many billions of molecules at one time and be done with it. Quantum methods are linear instead of cyclic.
Furthermore, high performance computing (HPC) is not inexpensive. Supercomputers, even with many graphics processing units (GPU), are slower, power hungry, environmentally unfriendly, and more expensive all at the same time. Drug discovery companies can at least speed them up somewhat by adding more nodes, but one estimate is that it will cost roughly 1,000 times more in order to do that as compared to using either quantum annealing or digital annealing. In fact, one estimate is that it would cost roughly $\$$40,000 to do what a quantum annealer could do for around $\$$40.
Quantum technology’s transformation of biopharmaceutical research and development, therefore, is threefold. First, larger problems can be solved with quantum technologies than would otherwise be the case. Second, new drugs can be discovered faster than ever before. And, third, quantum technologies are in multiple ways less expensive to use. They not only cost less to use, they also need to be used less frequently. Quantum computing and medicine are natural partners, and we can all hope that less expensive drug discovery may someday lead to less expensive pharmaceuticals for the world. And with time, all three of these transformations will continue to improve with future advancements in quantum computing.
How can quantum computing be used in medicine today?
Drug discovery firms such as Polaris Quantum Biotech are using quantum computing for drug development now. You can listen to CEO Shahar Keinan discuss their particular endeavors with CMO of Classiq, Yuval Boger, on his show, The Qubit Guy’s Podcast, here, where they discuss how Polaris Quantum Biotech is working with paying customers now on fee-for-service and collaboration projects. Collaborators may, for example, provide expertise finding targets while Polaris Quantum Biotech provides expertise finding molecules.
Quantum computing in medicine can also be defensive at this time. UnitedHealth Group (UHG), and its technology arm Optum Technology, for example, are hedging against future developments with quantum computing. They are defensively developing for the future, using patents and publications to establish and protect intellectual property. They are also exploring the potential for quantum machine learning (QML) to provide an even greater computational advantage than current methods.
As quantum computers mature, Classiq is strategically poised to help drug discovery firms achieve computational advantage with chemistry, optimization, search, and other problems. As mentioned earlier, designing quantum circuits to solve meaningful problems is normally challenging and time-consuming, but it doesn’t have to be. Schedule a demo of our synthesis engine and observe the future of drug development today.
Drug discovery companies can leverage quantum computing today to find compounds to treat and cure diseases. In conjunction with other leading-edge technologies, such as cloud computing, artificial intelligence (AI), and machine learning (ML), they can fully realize the potential of quantum computing for drug discovery by designing new molecular drugs. Quantum computing is significantly faster, much more efficient, and remarkably less expensive than all known alternatives.
As it pertains to drug development, the answer to why is quantum different lies in how very large problems can be addressed as compared to classical computers and even supercomputers. We’ll actually be discussing two different quantum technologies that are both used for computation. Annealing-based quantum computation is in use today and enables drug discovery firms to search chemical spaces a million times larger than would otherwise be the case. Gate-based quantum computing, although not yet ready for commercial use, stands to become even more powerful. Adding just one fundamental information unit, called a quantum bit or qubit, for short, increases the computational workspace exponentially. In other words, adding just one qubit to a gate-based quantum computer doubles the size of the problem we can address. For example, a small 40-qubit device can address a trillion different combinations. And what we’ll basically be discussing here, is searching a very large chemical space of molecules to find the select few that meet some specific criteria. Quantum computers handle large problems naturally.
How can quantum computing help drug development?
Beginning with a specific disease, or perhaps just a specific protein target of that disease, drug discovery companies can already start establishing the profile of a future drug. This drug profile will eventually take into account all the desired properties of this drug, including such considerations as specific protein binding sites, how the drug will enter the body (eg., injection, pill, etc.), how the drug may affect the brain, how much of the drug should be given, what is the target age group, compatibility with other drugs, schedule and duration of administration, packaging and shipping requirements, special marketing considerations, and so forth.
That drug profile is then used to generate a large virtual chemical space, which is when we can then start to see why there is an interest in quantum computing in the pharmaceutical industry. By “large,” we mean that the chemical space includes literally billions of molecules. The vast majority of those billions of molecules, however, are irrelevant to the selected drug profile. Therefore, quantum algorithms such as the Quadratic Unconstrained Binary Optimization (QUBO) algorithm, which is used for combinatorial optimization problems, can search this large chemical space for the specifically selected properties of the drug profile. The results of the QUBO algorithm are relevant to the drug profile in regard to both the desired properties and the desired binding site(s).
By greatly reducing the chemical space, quantum computing significantly reduces the extent of subsequent bench work that needs to be done. Bench work for drug discovery can include tests for toxicity, appropriate dosages, potential costs, and so forth. Completely inappropriate molecules have been removed from the chemical space. Far fewer molecules need to be designed, synthesized, measured, and experimented with. The path to defining and licensing assets is shortened. According to the University of Cincinnati College of Medicine, “a Phase I trial [currently] takes several months to complete” and only “about 70 percent of experimental drugs pass this initial phase of testing.” Decreasing the first statistic and increasing the second statistic presents a clear quantum computational advantage.
How will quantum computing affect medicine?
Quantum computing empowers altruistic goals, such as wanting to treat more diseases in less time than we could possibly do just a few decades ago, and maybe even just a few years ago. Target diseases might currently be of personal interest to researchers, or they might fill in some of the gaps in the present pharmaceutical market. One research goal might be to improve drug efficacy in an area of the market where the existing drugs fall short. Another research goal might be to find smaller molecules with which to improve delivery methods, perhaps allowing patients of all ages to put away their syringes and to start switching to ingesting digestible tablets or capsules instead. Quantum computing allows drug discovery firms to start building their asset portfolios right now with molecules that hold great potential to become life-changing medicines in the near future. Even getting these drugs just into the early stages of their development can shorten the time to put them into production later.
With quantum computing, pharmaceuticals can be brought to market much faster. The actual time it takes to reduce the initial chemical space, from many billions of molecules to mere hundreds of thousands of molecules, takes less than one minute. Of course, there’s much more to a project than a one-minute calculation, though. The quantum algorithm does require preparation -- designing quantum circuits at even small scales is remarkably challenging and time-consuming -- and then there is classical post-processing, as well. Traditional chemistry tools reduce the intermediate chemical space from hundreds of thousands of molecules, perhaps, to mere hundreds of molecules. Instead of drug discovery companies testing many billions of molecules, only the few hundreds of molecules that made it through both quantum and traditional processes need to be synthesized and measured. The project time might be around three months.
It is worth clarifying with a second mention that with the current state of quantum computing, that this is a two-step process. First, a quantum annealer -- a quantum technology that specifically addresses optimization problems -- reduces the initial chemical space from many billions of molecules to hundreds of thousands of molecules. The aforementioned QUBO algorithm runs on this quantum annealer. A distinct second step is needed to reduce this intermediate chemical space down to the final few hundreds of molecules, and this process still uses traditional chemistry tools. Although hundreds of molecules is a broad chemical space, those molecules at least have all the desired properties required by the selected drug profile. This hybrid quantum-traditional approach offers a significant computational advantage over a traditional-only approach, while leaving room for down the road when fully-quantum approaches will become available and the computational advantage will be even more pronounced. Hybrid approaches, besides leveraging quantum computing, usually incorporate other advanced technologies, such as cloud computing, artificial intelligence (AI), and machine learning (ML).
Generally speaking, drug discovery firms are not limiting their searches to molecules with specific properties for specific protein targets, although that is the most important outcome. As alluded to before with a mention of various delivery methods, these companies are also searching for the smallest possible molecules that can fit each drug profile. It’s very difficult, from the outset, to search a chemical space with literally billions of molecules for the desired drug properties. It’s even more difficult, as you may imagine, to find small molecules in such a large space. Like putting bumpers into bowling alley gutters, quantum computing makes it easier for drug discovery companies to find their targets: the smallest possible molecules that have all the desired properties of a selected drug profile.
It’s worth noting that one of the natural advantages of quantum computing is simulating quantum systems. Besides searching a chemical space for molecules that meet specific criteria, quantum computers can simulate the molecules themselves. One potential application of this is repurposing existing therapeutics. One affect on medicine could be skipping new drug development altogether, and repurposing drugs that have already been developed and for which at least some clinical data already exists.
What is the future of drug development?
According to the University of Cincinnati College of Medicine, phase 1, 2, and 3 clinical trials currently take several years, and only about 25-30% of drugs pass all three phases. Each phase tests the efficacy and safety of a drug, but increasing the sample population with each subsequent phase. Phase 1 trials may have less than 100 subjects, phase 2 trials may increase this to several hundred subjects, and phase 3 trials may include thousands of subjects. According to a JAMA paper, the average cost to develop a drug is just under US$\$$1B, with high-end estimates near US$\$$3B. The future of drug development is shorter clinical trials, greater effectiveness, and enhanced safety at a much lower cost.
The method that has been discussed in this blog article thus far uses a quantum technology called quantum annealing. There are actually several quantum annealer providers in the quantum technologies industry, at least three, in fact, that are well-known, and there is also such a thing as a digital annealer. A digital annealer, which could also be called “quantum-inspired,” can solve the same problems as quantum annealing despite not actually being a quantum technology. The primary advantage of both quantum annealing and quantum-inspired annealing is that they are large enough today to solve problems even when chemical spaces include the many billions of molecules that have been referred to in this article. In addition to that, there are presently-known methods to use existing quantum annealing and quantum-inspired annealing devices to work with much larger chemical spaces, if and when needed.
Another quantum technology called gate-based quantum computers, or universal quantum computers, are neither large enough nor accurate enough to be useful today. Unfortunately, they are still in their very early stages of development. However, they are most likely the future of drug development. Like quantum annealing and quantum-inspired annealing, they will be able to reduce large chemical spaces from their initial billions of molecules down to hundreds of thousands of molecules. But, unlike quantum annealing and quantum-inspired annealing, gate-based universal quantum computers will also be able to perform the computational chemistry calculations to further reduce the intermediate chemical space from hundreds of thousands of molecules to only hundreds of molecules. Solving Schrodinger’s equation with gate-based universal quantum computers will provide a computational advantage even over current hybrid quantum-traditional methods.
Reducing the overall time of new drugs to market, first with quantum annealers and digital annealers and later with gate-based universal quantum computers, is critical in this day and age. COVID therapies and vaccines were developed remarkably fast, and yet this historical timeframe remains relatively slow when compared to the quantum methods that we anticipate having in the near future. There is also ongoing concern about these therapies and vaccines, including in regard to their short-term efficacy and their long-term side effects. One of the many promises of quantum computing is discovering fast, safe, and effective medicines for all diseases, including in humanity’s fights against future pandemics. This may be through new drug development or, as noted earlier, through repurposing existing drugs.
Will quantum computing Transform Biopharma R&D?
Disregarding, for a moment, the future capabilities of gate-based universal quantum computers, quantum annealers and quantum-inspired digital annealers are transforming biopharmaceutical research and development in the present day. New drugs can already be discovered faster than they could ever be discovered with traditional methods. Plus, larger problems can be addressed.
Overall, the method of drug discovery is improved. Traditionally, a laboratory may identify thousands of molecules, for example. First note that this example uses a count in the thousands instead of in the billions. Now imagine that the desired properties are the same and the target protein is the same as if using either quantum annealing or quantum-inspired annealing. Yes, the traditional methods work. With time, the chemicals can all be synthesized and measured and tested. New drugs can be found, of course. However, imagine wanting to improve some of the drug’s properties. Or, imagine wanting to add some new properties to the drug. The traditional process becomes a cycle. Drug discovery companies design, synthesize, and measure initially, but desiring improvements of any kind require them to do it all over again.
With quantum annealing and digital annealing, and in the future with gate-based universal quantum computing, this drug development cycle ends. Instead of processing hundreds or even thousands of molecules at a time only to repeat the process for every desired improvement, drug discovery firms can process many billions of molecules at one time and be done with it. Quantum methods are linear instead of cyclic.
Furthermore, high performance computing (HPC) is not inexpensive. Supercomputers, even with many graphics processing units (GPU), are slower, power hungry, environmentally unfriendly, and more expensive all at the same time. Drug discovery companies can at least speed them up somewhat by adding more nodes, but one estimate is that it will cost roughly 1,000 times more in order to do that as compared to using either quantum annealing or digital annealing. In fact, one estimate is that it would cost roughly $\$$40,000 to do what a quantum annealer could do for around $\$$40.
Quantum technology’s transformation of biopharmaceutical research and development, therefore, is threefold. First, larger problems can be solved with quantum technologies than would otherwise be the case. Second, new drugs can be discovered faster than ever before. And, third, quantum technologies are in multiple ways less expensive to use. They not only cost less to use, they also need to be used less frequently. Quantum computing and medicine are natural partners, and we can all hope that less expensive drug discovery may someday lead to less expensive pharmaceuticals for the world. And with time, all three of these transformations will continue to improve with future advancements in quantum computing.
How can quantum computing be used in medicine today?
Drug discovery firms such as Polaris Quantum Biotech are using quantum computing for drug development now. You can listen to CEO Shahar Keinan discuss their particular endeavors with CMO of Classiq, Yuval Boger, on his show, The Qubit Guy’s Podcast, here, where they discuss how Polaris Quantum Biotech is working with paying customers now on fee-for-service and collaboration projects. Collaborators may, for example, provide expertise finding targets while Polaris Quantum Biotech provides expertise finding molecules.
Quantum computing in medicine can also be defensive at this time. UnitedHealth Group (UHG), and its technology arm Optum Technology, for example, are hedging against future developments with quantum computing. They are defensively developing for the future, using patents and publications to establish and protect intellectual property. They are also exploring the potential for quantum machine learning (QML) to provide an even greater computational advantage than current methods.
As quantum computers mature, Classiq is strategically poised to help drug discovery firms achieve computational advantage with chemistry, optimization, search, and other problems. As mentioned earlier, designing quantum circuits to solve meaningful problems is normally challenging and time-consuming, but it doesn’t have to be. Schedule a demo of our synthesis engine and observe the future of drug development today.
About "The Qubit Guy's Podcast"
Hosted by The Qubit Guy (Yuval Boger, our Chief Marketing Officer), the podcast hosts thought leaders in quantum computing to discuss business and technical questions that impact the quantum computing ecosystem. Our guests provide interesting insights about quantum computer software and algorithm, quantum computer hardware, key applications for quantum computing, market studies of the quantum industry and more.
If you would like to suggest a guest for the podcast, please contact us.
