PubMed · 2026-08-04
AI can now predict protein shapes with impressive speed, but plants have signaling systems, hormone receptors, and immune proteins that only reveal how they actually work when scientists look at them directly in the lab. This review argues that real experiments, not just computer predictions, are still essential for understanding how plants sense their environment and defend themselves.
Plant signaling proteins often involve dynamic processes like ligand-triggered shape changes and transient interactions that static AI predictions can't capture
Plant proteins are markedly underrepresented in structural databases compared to other organisms, leaving many plant-specific pathways unresolved
Historical examples like the first crystallization of urease and recent cryo-EM studies of immune and hormone receptor complexes show experimental structural biology repeatedly uncovers mechanisms AI alone would miss