Spotlight on AI and Scientific Discovery

This edition explores how artificial intelligence is reshaping scientific discovery across molecular prediction, biological mapping, and model transparency. From structure-based foundation models that predict protein folds, ligand binding, energetics, and conformational change, to proteome-scale inference of human protein-protein interactions, these advances are turning computation into a powerful engine for decoding cellular function. Generative deep learning further expands this shift by emulating protein ensembles and free-energy landscapes at speeds far beyond conventional simulations. Alongside these AI-driven tools, the collection highlights the value of openness and scrutiny in model development, as illustrated by the analysis of DeepSeek R1, and the continued importance of high-resolution experimental resources such as REPAIRome and ExIGS. Together, these studies show that discovery increasingly emerges from the interplay between predictive models, scalable datasets, and experimentally grounded biological insight.