AI x INTENDD® – prediction of activity and beyond –
Small molecules (PPI modulators) designed by AI-guided INTENDD® will not require further activity optimization
AI-guided INTENDD® drastically improves drug discovery productivity of hit identification to lead optimization by predicting the binding pose of small molecule to target protein and separating its activity into 8 classes of 3-digits pmol/L to 1-digit mmol/L orders.
Interprotein conducts the development of INTENDD-based AI drug discovery technology in collaboration with A.I. Squared, Inc., which has an excellent deep learning technology, and has established AI-guided INTENDD®, an activity prediction system for small molecules. The activity prediction by this system reaches a practical level with the highest accuracy all over the world and is expected to change the paradigm of current drug discovery processes. We are aiming at AI-based de novo drug design without dependence on compound libraries as a final goal.
Separation of activity into 8 classes by AI-guided INTENDD® enables to eliminate the need for the enormous cycle of compound synthesis and evaluation in the conventional drug discovery processes. In an improved process by AI-guided INTENDD®, medicinal chemists can lean activity of designed compounds without those synthesis and evaluation, and for example it is only necessary to synthesize the highly active compounds that are separated into Class 0 to 2. Such approach is expected to contribute to the efficient evaluation for selectivity, toxicity, DMPK and so on.