Worth knowing
Controlling Biology - Fabian Theis, Helmholtz Munich
It provides insight into the methodological bottlenecks in AI for biology and why major institutions are investing billions in new multi-modal data collection to make virtual cell models viable.
Reported by Latent Space ·
What happened
Latent Space features an interview with Fabian Theis of Helmholtz Munich discussing the challenges and future of AI-driven Virtual Cell modeling. The post examines why complex models have struggled against simple baselines in perturbation prediction and how a recent $1.8 billion data collection effort involving Biohub, NIH, DOE, Google, and Meta aims to advance multi-scale biological modeling.
Who it’s about
- Fabian Theis
- Helmholtz Munich
- Biohub
- Meta
Source
Latent Space
Controlling Biology - Fabian Theis, Helmholtz Munich