Recently, scientists from China and the United States introduced a new artificial intelligence model called ActFound, which promises to revolutionize the drug discovery process. This AI model, combining meta-learning and pairwise learning, is capable of overcoming traditional challenges in drug development, such as limited data sets and incompatible measurements. With these innovations, ActFound can significantly accelerate the creation of new drugs while reducing costs.
ActFound was trained on a vast amount of data: over 35,000 assays and 1.6 million measured biological activities from a leading chemical database. This allowed the model to outperform nine competing counterparts and demonstrate high accuracy in predicting biological activity, especially for anticancer drugs. These results suggest that ActFound could become a key tool in speeding up the discovery of new drugs, which is particularly crucial for combating diseases like cancer.
In recent years, AI has been actively integrated into China’s pharmaceutical industry, where the country’s infrastructure and technological capabilities facilitate fast and precise research. It is predicted that in the coming years, the use of such technologies will help significantly reduce risks in the development of new drugs and enhance their effectiveness.
