Introduction

SandboxAQ has announced a collaboration with Nvidia to combine its large quantitative models (LQMs) with the Nvidia CUDA-accelerated Density Matrix Renormalization Group (DMRG) algorithm. This partnership aims to perform highly accurate quantitative AI simulations of real-life systems, surpassing the capabilities of traditional AI models.

Enhanced Computational Power

According to SandboxAQ, the integration of these technologies will enable computing speeds exceeding 80x compared to traditional 128-core CPUs. Nadia Harhen, general manager for SandboxAQ’s AI simulation division, highlighted that this advancement supports computational chemistry applications across various industries, including biopharma, chemicals, and materials sciences.

Equation-Based Computing

The technology employs complex physics-based calculations to generate data necessary for training SandboxAQ’s LQMs. These models can predict molecular behavior and create molecular structures with desired properties. Harhen stated, “Equation-based computing is a highly efficient way of programming computers,” which facilitates the development of applications tailored to drug discovery needs.

Applications in Drug Discovery

One notable tool is AQ-FEP, an absolute free energy perturbation model that ranks potential drug molecules by their likelihood of success. Harhen emphasized the need for complex models in drug discovery, stating, “We have to invent new things that aren’t close to the training dataset. And this is where you actually have to ground the AI in the laws of physics.”

Real-World Impact

The collaboration allows SandboxAQ to perform challenging computations more rapidly. For instance, CUDA-DMRG can accurately model the catalytic activity of Cytochrome P450, crucial for drug metabolism and toxicity. The partners have published a pre-print that shares results from applying the DMRG algorithm on Nvidia hardware.

Future Directions

SandboxAQ is also exploring catalysts that can penetrate the blood-brain barrier, collaborating with scientists at the University of California, San Francisco’s Institute for Neurodegenerative Diseases to identify new treatments for disorders like Parkinson’s and Alzheimer’s diseases.

Industry Interest

SandboxAQ’s technology has attracted attention from major pharmaceutical players, with collaborations announced with AstraZeneca and Sanofi. The partnership with Nvidia enhances the appeal of SandboxAQ’s offerings, as more biopharma companies seek to leverage AI technologies for accelerated drug discovery.

Conclusion

As AI continues to evolve, the need for advanced solutions in drug discovery becomes increasingly apparent. Harhen noted, “Many companies are making incremental changes using AI and they are having breakthroughs,” but emphasized the importance of a different kind of AI to tackle challenging diseases effectively.