One of the first and most promising uses scientists envision for the rapidly evolving technology of quantum computing is a new approach to drug development. A quantum computer could, in theory, eliminate much of the trial and error involved in the process to help researchers more quickly zero in on ways to treat aggressive cancers, prevent dementia, kill deadly viruses or even slow aging by sifting through the trillions of molecules that might potentially be synthesized to create pharmaceuticals.
As proof of the technology’s potential, a group of researchers published a paper in Nature Biotechnology earlier this year showing how they could use a small-scale quantum computer designed by IBM and AI to identify a potential cancer drug.
While several dozen quantum computers are working in labs worldwide, they’re not yet advanced enough or big enough to beat existing supercomputers except for certain special test problems. Still, there have been some surprising leaps in progress.
“We’re not making the claim that it’s faster, cheaper, better or anything … we’re showing it’s possible,” said Alex Zhavoronkov, a co-author of the paper and founder of Insilico Medicine. He compares these early uses with the first airplane flights — essential for demonstrating a new mode of transportation once deemed impossible.
Until recently, quantum computers were severely limited by their tendency to make errors. They use units of information storage called qubits, and stringing them together only compounds the error rate. Last year, the startup Quantinuum and later Google announced they’d found a way to resolve the problem so that adding more qubits decreased the error rate by building in a kind of redundancy.
While ordinary computers store information in bits, which can take the values 0 or 1, a qubit can take on both values simultaneously, enabling quantum computers to process data in fundamentally different and often more powerful ways.
Quantum computing harnesses the famously strange behavior of quantum physics, where atoms, light and subatomic particles exist in states of uncertainty until observed — even their position can resemble a smeared-out wave rather than a single point in space.
Qubits can be created in various ways — from electrons moving through supercooled materials to atoms suspended in place by lasers. Most current systems connect only a handful of qubits, but Google set a milestone last December by implementing error correction in a system of 105. If this approach can be scaled to thousands of them, scientists believe it could revolutionize how we tackle real-world complexity — enabling breakthroughs in medicine, energy storage, high-efficiency solar panels, next-generation space suits, and innovations we haven’t yet imagined.
It’s exciting how quickly the field is advancing, said Brian DeMarco, a physicist who studies quantum computing at the University of Illinois, Urbana-Champaign.
In DeMarco’s lab, researchers make qubits from the spin of single atoms. He said these atoms can be isolated from their environment so well that their quantum behavior dominates, enabling them to be used as qubits for quantum computing.
The scientists involved in the cancer drug research used a system with just 16 qubits to find a new molecule capable of binding to a protein called KRAS. The protein has proved hard to target with existing drugs.
Christoph Gorgulla, a biologist at St. Jude Children’s Research Hospital in Memphis, said the researchers eventually hope to be able to specify an action for a drug to carry out and then use quantum computers to search for the right molecules for the job. He said the number of drugs that could potentially be developed through this process could be described as 10, followed by about 60 zeros.
It’s not so much that the quantum computer is fast, he said, but it speaks the language of matter, so it takes fewer steps to get to the same place. “On this atomic level, it’s really quantum mechanics that governs what is happening … how the atoms move, how they interact, and how strongly,” said Gorgulla, one of the study’s co-authors.
DeMarco agreed. “The reason that protons and neutrons and electrons can arrange themselves into atoms is because of quantum physics,” he said. He said the rules of chemistry are sometimes enough, but often, they fall short. Quantum physics offers a master formula — the Schrödinger equation — for predicting how matter behaves. The problem is that it’s unusable for the complex molecules that make up our bodies; solving it with conventional computers would take millions of years.
Scientists are reluctant to predict precisely when quantum computers will be capable of speeding the discovery of drugs, chemicals and new materials, but many envision it happening within a decade. Last month, DARPA launched its “Quantum Benchmarking Initiative,” aiming to chart a path toward an industrially viable quantum computer by 2033.
More research is needed to continue progressing in the field and for the US to maintain its place in the race. Last week, several of the industry’s leaders appeared before Congress to advocate for continued government support.
Michael Kratsios, President Donald Trump’s science adviser, has championed quantum computing and AI. However, there are concerns that the administration’s budget cuts — especially in research — will set efforts back. The drastic cuts have already led some scientists to work elsewhere. The Nature Biotechnology paper’s lead author, physicist Alán Aspuru-Guzik, left Harvard for the University of Toronto following Trump’s first election in 2016, citing concerns about the country’s political climate.
Uncertainty is part of the nature of science — we can’t always predict where a pursuit will lead or how long it will take to produce practical results. One thing we can predict is that giving up guarantees we’ll fall behind.
A message from Advisor Perspectives and VettaFi: To learn more about this and other topics, check out some of our webcasts.
Bloomberg News provided this article. For more articles like this please visit
bloomberg.com.
Read more articles by F.D. Flam
