Activity 2
Generative drug design
Virtual exploration of ~1023 – 1060 molecules in unexplored, underexploited chemical space to identify novel antimicrobials
Activity leader: Audrey Durand
Anticipated output
Predictive AI models for target binding / antibiotic activity
Generative AI models for antibiotic drug design
Virtual compound libraries for experimental testing
Platform development
Computation: AI-based drug design, data analysis and integration
Compute hosted by U. Laval
Chemical space is vast and largely unexplored
AI can help identify novel classes of antibiotics
To overcome existing mechanisms of AMR, it is important to discover new classes of antibiotics:
Interactive learning iteratively improves AI-based drug design
