Chinese Researchers Unveil AI Platform DrugCLIP That Accelerates Drug Screening by Million-Fold
核心洞察
Chinese researchers at Tsinghua University have developed DrugCLIP, an AI-powered virtual drug screening platform that achieves a million-fold increase in screening speeds compared to conventional molecular docking methods.
The platform enabled the first-ever virtual screening project on a human-genome scale, analyzing over 500 million drug-like molecules across 10,000 protein targets (搜索) and yielding more than 2 million potential active compounds.
DrugCLIP compresses what would take centuries of computation into a single day on one computing node, creating the largest known protein-ligand screening database that is freely available to the global scientific community.
Chinese researchers at Tsinghua University have developed an AI-powered virtual drug screening platform called DrugCLIP that achieves a million-fold increase in screening speeds over conventional methods, according to a study published in Science. The breakthrough technology has enabled the first-ever virtual screening project on a human-genome scale, opening new avenues for innovative drug discovery.
Addressing a Critical Bottleneck in Drug Discovery
The human genome encodes over 20,000 proteins, yet only a small fraction have been explored as drug targets. Currently, the vast majority of the human genome's potential targets and compounds remain untapped, creating a major bottleneck in drug discovery. Screening for promising lead compounds across this vast chemical and target space has severely limited progress in the field.
Traditional molecular docking approaches face significant computational limitations. Screening just 10,000 protein targets (搜索) using state-of-the-art molecular docking tools would take centuries of continuous computation on a single computer, severely hindering the matching of novel targets with potential molecules.
Revolutionary Speed Enhancement
DrugCLIP addresses this challenge by dramatically reducing screening timeframes, compressing what would take centuries into a single day on one computing node. The platform's breakthrough lies in reformulating traditional molecular docking as a high-efficiency semantic search for protein pockets (搜索) and small molecules in vector space.
Running on a computing node with a 128-core CPU and 8 GPUs, DrugCLIP can score trillions of protein pocket-small molecule pairs daily, delivering the million-fold speed increase over conventional docking approaches. This represents a fundamental advancement in computational drug discovery capabilities.
Genome-Scale Screening Achievement
In its inaugural genome-scale screening project, the research team applied DrugCLIP to approximately 10,000 protein targets (搜索) and 20,000 protein pockets (搜索), analyzing over 500 million small, drug-like molecules. The comprehensive effort yielded more than 2 million potential active molecules, resulting in the largest known protein-ligand screening database to date.
This database is now freely available to the global scientific community, offering powerful data support for fundamental research and early-stage drug discovery. The open access approach aims to accelerate collaborative efforts in drug development worldwide.
Platform Adoption and Service Implementation
A companion screening service platform has been launched alongside the database, allowing users to submit custom targets and protein pockets (搜索) for analysis. As of the study's publication, the platform has demonstrated significant adoption, serving over 1,400 users and completing more than 13,500 screening tasks within a span of six months.
Complementary Advancement to AlphaFold
While AlphaFold, which won the 2024 Nobel Prize in Chemistry, addressed the challenge of protein structure prediction, DrugCLIP makes a complementary leap by establishing a critical bridge from protein structure to drug discovery. This connection enables large-scale virtual screening across the human genome, building upon the structural insights provided by AlphaFold.
Future Therapeutic Applications
The DrugCLIP team plans to collaborate closely with academic and industry partners to accelerate the discovery of novel targets and first-in-class therapies in areas such as oncology (搜索), infectious diseases (搜索), and rare disorders (搜索). The researchers will continue to enhance the platform's performance, expand its capabilities, and contribute to a more intelligent, efficient, and accessible global ecosystem for drug innovation.