Structure-based AI design of protacs for drug discovery

Overview

The proposed solution utilizes advanced AI-driven, structure-based design methods to generate proteolysis targeting chimeras (PROTACs) with optimized chemical properties for drug discovery. By integrating biophysics simulations, this approach ensures the creation of compounds that effectively target and degrade proteins within living cells. The solution is particularly valuable for designing PROTACs that enhance interactions with neosubstrates and improve target selectivity, offering significant advancements in therapeutic applications.

Technical specifications

Key features:

• AI Platforms:
  • Platform 1: MERES for goal-directed molecular design.
  • Platform 2: Predicts protein conformational changes upon ligand binding.
  • Platform 3: Accelerates conformational sampling with a diffusion AI model.
  • Platform 4: Predicts weak protein-protein interactions using a transformer AI model.
• Design Process:
  • Step 1: Binder design using MERES, requiring a target structure and starting molecule.
  • Step 2: Ternary structure modeling, utilizing candidate structures and E3/POI inputs.
  • Step 3: Linker engineering, generating linkers that fit the geometry of the ternary structure.
  • Step 4: Prioritization of candidates through energy scoring and cross-validation.

Technology readiness level

This research is currently at Technology Readiness Level 4, indicating that the technology has been validated in a laboratory setting and is ready for further development and testing in more practical environments.