2228 JingTai Technology Visit Records and Thoughts 02.4

Technical Capabilities and Moat:

Has JingTai developed AI algorithms with independent intellectual property rights? What advantages do these algorithms have in handling complex molecular structures and biological processes?

The company's quantum AI technology has surpassed the efficiency of pure AI solutions. The company achieves a full R&D process from dry lab calculations and evaluations to wet lab experiments through first-principles calculations based on quantum physics, artificial intelligence, and automation. The AI methods based on quantum physics have advantages over traditional manual methods and AI-based methods in terms of speed, scale, novelty, and preclinical development success rates in drug discovery.

The exceptional computational power of quantum physics enables the company to accurately predict and simulate the structures and properties of molecules, thereby efficiently identifying and optimizing candidate compounds in the R&D of new drugs and materials science. The company provides atomic and electronic-level mechanistic explanations centered on first principles, driving innovative research. Quantum physics-based calculations can improve the R&D cycle through high-precision predictions and descriptions, saving time and costs, and enhancing data-driven capabilities.

Leveraging the advantages of quantum AI technology, the company has launched three tools: XFF, XFEP, and XPose, which can be customized for different application scenarios, providing efficient and accurate exploration and development for the key chemical space of new drugs and materials. XFF is a general molecular force field for global optimization of parameter design, co-developed with Pfizer from 2018 to 2021, capable of comprehensively covering the chemical space of new drugs and materials, accurately describing and predicting properties, and flexibly deploying and customizing adaptation parameters; XFEP is an AI-powered high-precision and high-throughput binding affinity prediction platform, providing a foundation for accurate, all-scenario, efficient, scalable, and cost-effective FEP applications in drug and materials science R&D; XPose is an internal binding mode prediction platform that can simulate the free energy of drug or target complexes to predict the binding conformations of small molecule targets-ligands, and can also be used to construct precise SAR and structure-based affinity assessments, bridging the gap between predicted target structures and the practical applications of different drug discovery scenarios.