We employ a virtual-physical strategy that integrates parallel approaches for efficient ultra-high-throughput screening.

The foundation of our screening system is a massive library containing tens of billions of virtual and physical compounds. We have built this asset by combining our curated internal resources and those from leading academic and industrial partners.

The library comprises

a high-quality in-house combinatorial chemistry library built by generative AI models

an industry-leading DNA-encoded library introduced through collaboration

a physical high-throughput screening library co-developed with Westlake University and Westlake Laboratory.

Building upon the extensive compound library, we have developed a framework for hit identification and validation. It integrates a suite of advanced technologies, including AI prediction, high-precision computing, physicochemical determination, and automated biochemical assays. This system enables systematic and precise screening of ultra‑large‑scale compound libraries, speeding up hit discovery.