A new AI breakthrough could slash the time it takes to discover life-saving medicines, according to researchers in China.
Scientists at Tsinghua University have developed a powerful system called DrugCLIP, which can screen drug molecules against human proteins at staggering speed compared to traditional drug-testing methods.
According to Physics.org, DrugCLIP uses deep contrastive learning to convert both drug molecules and protein binding pockets into digital vectors, allowing the system to match them almost instantly.
The AI screened 500 million molecules across 10,000 human proteins in testing, covering around half of the human druggable proteome.
Researchers say the system carried out 10 trillion molecule–protein evaluations in a single day, making it around 10 million times faster than classic docking simulations, which are commonly used in early drug discovery.
To build the platform, the team used AlphaFold2 to generate protein structures and refined the binding sites with a custom tool called GenPack. The results were then validated using both computer modelling and laboratory experiments.
In their paper, the scientists said: “DrugCLIP is an ultrafast virtual screening method that we rigorously validated through in silico benchmark evaluation and wet-lab experiments.”
They added, “Its speed enables trillion-scale screening covering the human druggable proteome, providing an open-access resource that forms a foundation for next-generation drug discovery, particularly for less understood targets.”
Notably, the AI identified potential compounds for TRIP12, a protein linked to cancer and autism that has so far resisted traditional drug-targeting efforts.
All of DrugCLIP’s data and models are freely available, meaning laboratories around the world can now use the system to dramatically speed up early-stage drug development.
