The study was published May 13 in the American Chemical Society Chemical Media JournalInsilico Medicine (“Insilico”), a clinical-stage artificial intelligence (AI) drug discovery company, announced that it is combining two rapidly developing technologies, quantum computing and generative artificial intelligence, to explore leading drug development candidate discovery and successfully demonstrate potential advantages. Quantitative generative adversarial networks in generative chemistry.ation and modeling A leading journal of computational modeling, it was led by Insilico Centers in Taiwan and the UAE that focus on pioneering and building cutting-edge methods and engines with rapidly evolving technologies – including generative AI and quantum computing – to accelerate drug discovery and development. The research was supported by University of Toronto Acceleration Consortium director Alan Aspuru-Juzek, Ph.D., and scientists from the Hon Hai Research Institute (Foxconn).
This international collaboration was a very interesting project. It paves the way for further developments in artificial intelligence as it meets drug discovery. This is a global collaboration where Foxconn, Insilico, Zapata Computing and the University of Toronto are working together.”
Alán Aspuru-Guzik, director of the Accelerometer Consortium and professor of computer science and chemistry at the University of Toronto
Generative adversarial networks (GANs) are one of the most successful generative models in drug discovery and design and have shown remarkable results for generating data that simulates data distribution in different tasks. The classic GAN model consists of a generator and a discriminator. The generator takes random noise as input and tries to mimic the distribution of the data, and the discriminator tries to distinguish between fake and real samples. The GAN is trained so that the discriminator cannot distinguish generated data from real data.
In this paper, the researchers explored the quantum advantage in small-molecule drug discovery by replacing each part of MolGAN, an implicit GAN for small-molecule graphs, with a variable quantum circuit (VQC), step by step, including a noise generator, generated using the correction method, and a quantitative discriminant, comparing its performance with its classical counterpart.
Not only did the study demonstrate that trained quantum GANs can generate molecules similar to the training set using VQC as a noise generator, but the quantum generator is superior to classical GANs in the pharmacological properties of the generated compounds and the target-oriented benchmark. In addition, the study showed that a GAN quantum discriminant with only tens of learnable parameters can generate valid particles and outperforms the classical counterpart by tens of thousands of parameters in terms of the molecule properties generated and the degree of KL divergence.
Quantum computing has been recognized as the next technological breakthrough that will have a huge impact, and the pharmaceutical industry is believed to be among the first wave of industries to benefit from the advances. This research demonstrates Insilico’s first footprint in quantum computing using AI in molecular generation, confirming our insights in this field.”
Jimmy Yen-Chu Lin, PhD, general manager of Insilico Medicine Taiwan and corresponding author of the paper
Building on these findings, Insilico scientists plan to integrate a hybrid quantum GAN model into Chemistry42, the company’s proprietary small molecule generation engine, to further accelerate and optimize AI-driven drug discovery and development.
Insilico was one of the first to use GANs in de novo molecular design, publishing the first paper in the field in 2016. The company submitted 11 preclinical candidates through generative AI models based on GANs and its pilot program was validated in phase I clinical trials. .
“I am proud of the positive results our quantum computing team has achieved through their efforts and innovation,” said Alex Zhavoronkov, Ph.D., founder and CEO of Insilico Medicine. “I think this is the first small step in our journey. We are currently working on a breakthrough experiment with a real quantum computer for chemistry and look forward to sharing Insilico’s best practices with industry and academia.”