In a study published in PNAS, a research team developed a new reinforcement learning-based enhanced sampling method called Adaptive Collective Variables Generator (Adaptive CVgen), which has been successfully applied to study protein folding and the synthesis of fullerene (C60).
Source:
phys.org