The AI for Science Computing Laboratory is a cutting-edge research platform focusing on interdisciplinary innovation between artificial intelligence and basic sciences. It is committed to building an intelligent research system integrating "platform-model-technology" as a trinity. The laboratory sets its development goal as "creating three firsts": establishing China’s first provincial-level key laboratory for AI for Science computing and applications, building the first full-chain research platform connecting "education-technology-industry-talent" across AI and basic sciences, and creating Guangdong’s first AI4S industrial technology incubation base.

The laboratory focuses on four core research directions:

1.Basic theories and methods of AI for Science computing

It focuses on researching data-and-principle dual-driven scientific models, cross-scale characterization, and design optimization. It aims to establish basic theories and methods of AI for Science for the life sciences field, including developing multi-scale AI for Science large models for living organisms across temporal and spatial dimensions, building multi-modal AI for Science large models for various molecular modalities, and researching AI+HPC integrated computing methods.

2.AI-enhanced genome analysis and mining

Based on a self-built large-scale prokaryotic genome database, it adopts high-throughput annotation and AI pre-trained large models to identify operon sequences, enabling efficient and accurate recognition of protein complex coding sequences and regulatory genetic elements.

3.AI-assisted evolution of novel proteins

It builds a closed-loop system of "generation-screening-self-optimization-experimental verification-model update" centered on generative protein language models and molecular dynamics simulations, realizing generative design and optimization of drug lead compounds.

4.AI-accelerated innovative drug discovery

It integrates molecular dynamics into the training process of AI large models, acquires drug target mechanisms through large model-driven four-dimensional cryo-electron microscopy technology, realizes complex dynamics prediction of proteins and drug molecules, and assists in precise drug discovery.

With the feature of "intelligent computing driving scientific discovery," the laboratory conducts interdisciplinary scientific research using AI technologies, particularly tackling key technologies in AI-assisted genome analysis, protein evolution, and drug discovery. It strives to promote in-depth integration of artificial intelligence and scientific research, providing core driving force for cultivating new productive forces in AI for Science in the Guangdong-Hong Kong-Macau Greater Bay Area.