Teaching AI to Understand Biology: From Cellular Perturbations to Intelligent Drug Discovery

Thursday, January 28, 2027 at 3:10 PM to 3:40 PM · 30 min. (Europe/London)
AI in Drug Discovery Stage
Day 2 - Thursday 28th January
AIAI in Drug DiscoveryDrug Discovery

Information

  • Can AI predict cellular responses to drugs? — Drawing on MechPert, LangPert, and SynthPert, we explore how fusing mechanistic knowledge, language, and synthetic data leads to more accurate and scientifically meaningful perturbation predictions. 
  • Can LLMs truly reason about biology? — Moving beyond mere pattern recognition, we will examine how supervised and reinforcement learning can train models to tackle complex biological problems with verifiable, trustworthy logic. 
  • Can AI agents execute bioinformatics? — Sharing practical lessons from deploying autonomous agents at Novo Nordisk, we will discuss why rigorous evaluation, transparency, and uncertainty-awareness are critical to making AI a reliable partner in drug discovery.