The AI Enabled Superkine platform utilizes the latest self-training deep learning model to describe protein-protein interactions, enabling large-scale and efficient virtual screening of protein mutation combinations. This leads to improved stability of target proteins and increased affinity for targets. The results from the virtual screening are then verified through interactive computation chemistry methods, ensuring the accuracy of the design to the greatest extent. This approach streamlines the drug development process and saves on the cost of later experimental validation.
The AI-Enabled Superkine platform so far obtained twelve mutants as part of our new drug development pipeline through the first round of optimization design. Four of these mutants have been synthesized and verified to significantly outperform the unmutated drugs, and are currently undergoing further testing and validation.
