New research by Zapata Computing, Insilico Medicine, Foxconn, and the University of Toronto has shown the potential for quantum-enhanced generative models to outperform classical generative models in discovering small molecules with pharmaceutical value. The quantum-enhanced generative adversarial networks (GAN) generated small molecules with more desirable properties than those produced by purely classical GANs. This could help reduce time and costs for pharmaceutical research and development.
Quantum-Enhanced Generative Models for Drug Discovery
A recent study by Zapata Computing, Insilico Medicine, Foxconn, and the University of Toronto has demonstrated the potential of quantum-enhanced generative models to outperform classical generative models in discovering small molecules with pharmaceutical value. The research focused on hybrid quantum-classical generative adversarial networks (GAN) for small molecule discovery.
The teams replaced each element of GAN with a variational quantum circuit (VQC). They compared the molecules generated by the quantum-enhanced GANs with those caused by a purely classical GAN. The results showed that the small molecules created using a VQC frequently had better physicochemical properties and performance in the goal-directed benchmark than their classical counterparts.
Advantages of Quantum-Enhanced GANs in Drug Discovery
The drug discovery pipeline is traditionally a lengthy and costly process. However, recent advances in machine learning and deep learning technologies have proven to help reduce time and costs for pharmaceutical research and development. Using quantum-enhanced GANs, the researchers uncovered molecule designs with viable structures comparable to those from classical methods.
Alex Zhavoronkov, PhD, founder and CEO of Insilico Medicine, said that the collaboration with Zapata and Foxconn has demonstrated the potential of quantum and quantum-inspired generative models in transforming drug design and development through artificial intelligence.
Quantum Computing Applications in Other Industries
Quantum computing has the potential to solve complex computational problems in various industries. Zapata Computing has a track record of breakthrough research in quantum generative AI. In 2021, Zapata researchers were the first to generate high-resolution images using quantum generative models.
In more recent work with BMW, Zapata researchers demonstrated how quantum-inspired generative models could improve upon best-in-class traditional optimization solutions for a vehicle manufacturing scheduling problem. Yudong Cao, CTO and co-founder at Zapata Computing expressed excitement about the potential of quantum-inspired techniques to advance the pharmaceutical industry and other industries facing complex design challenges.
Zapata Computing builds solutions to enterprises’ most computationally complex problems. It has pioneered proprietary methods in generative AI, machine learning, and quantum techniques that run on classical hardware (CPUs, GPUs). Zapata’s Orquestra® platform supports developing and deploying better, faster, more cost-effective models, such as Large Language Models, Monte Carlo simulations, and other computationally intense solutions. Founded in 2017, Zapata is headquartered in Boston, Massachusetts.
For more information about the research conducted by Zapata Computing, Insilico Medicine, Foxconn, and the University of Toronto, visit www.pubs.acs.org/doi/full/10.1021/acs.jcim.3c00562.
“The drug discovery pipeline is traditionally a long and costly process, but recent advances in machine learning and deep learning technologies have proven to help reduce time and costs for pharmaceutical research and development. By working with Zapata and Foxconn, we uncovered molecule designs with viable structures comparable to those from classical methods.” – Alex Zhavoronkov, PhD, founder and CEO of Insilico Medicine.
“We are pleased to achieve this milestone in collaboration with Insilico Medicine. Quantum computing can be used to solve complex computational problems. The application of quantum computing in drug discovery will potentially help reduce the time and lower the cost of research and development,” – Min-Hsiu Hsieh, PhD, Director of the Quantum Computing Research Center of Hon Hai Technology Group (Foxconn).
“This work with Insilico Medicine and Foxconn is a great example of how quantum-enhanced generative AI can be used to solve real-world problems more effectively,” – Yudong Cao, CTO and co-founder at Zapata Computing.
Summary
Quantum-enhanced generative models have shown the potential to outperform classical models in discovering small molecules with pharmaceutical value. This breakthrough by Zapata Computing, Insilico Medicine, Foxconn, and the University of Toronto could help reduce time and costs for pharmaceutical research and development.
- Zapata Computing, Insilico Medicine, Foxconn, and the University of Toronto have collaborated on research exploring hybrid quantum-classical generative adversarial networks (GAN) for small molecule discovery.
- The teams used artificial intelligence and quantum computing techniques to replace each element of GAN with a variational quantum circuit (VQC).
- The molecules generated by the quantum-enhanced GANs had more desirable properties than those generated by purely classical GANs.
- The drug discovery pipeline could potentially be shortened and costs reduced through the use of quantum-enhanced generative AI.
- Zapata has previously demonstrated breakthrough research in quantum generative AI, including generating high-resolution images and improving vehicle manufacturing scheduling with BMW.
- Zapata’s Orquestra platform supports the development and deployment of generative AI, machine learning, and quantum techniques on classical hardware.
