Fabian Theis on where biological prediction should go

On November 6, 2026, Latent Space published an interview with Fabian Theis of Helmholtz Munich on the direction of biological prediction. Theis argues that perturbation prediction built on existing datasets cannot support complex algorithms. Research should instead use large-scale, high-precision, multi-scale registered biological data to predict evolution from molecules and cells up through tissues, organs and organisms. The piece gives his view of current data limits and future directions; it does not mention implementation progress or validation results for specific projects.

Written by AI from the coverage · updated 2 hours ago

Not enough continuous observations to draw a trend yet.