Accelerating Discovery with AI

The discovery phase is long and often unsuccessful. To speed things up, AI is now being used to make parts of drug discovery faster and more efficient.

AI has shortened the process from 10-15 pre-clinical years to 1 year.

There are many areas where AI and computation play a key role, including:

• Protein structure prediction, using tools like AlphaFold (note: UVA HPC does not currently have access) and RosettaFold.

• Kinetic predictions, which involve molecular simulations to understand how drugs interact over time.

• Peptide design, focusing on simulating interactions between proteins and peptides.

• Small-molecule design, aimed at discovering and optimizing drug-like compounds.

• Antibody-oriented design, used to develop targeted biologic therapies.

• PROTACs (Proteolysis Targeting Chimeras), a newer area focused on degrading disease-causing proteins.

These examples are just the beginning—there are many more computational tools and methods actively transforming the landscape of drug discovery.

This workshop will focus on small-molecule–based therapeutics. However, it’s important to remember that all areas of drug discovery are interconnected.