Recent advances in AI-driven biotherapeutic discovery using protein generative models

Protein generative models design therapeutic backbones and sequences that are then filtered by cross-validation metrics. We review these methods, their therapeutic applications, and how far these metrics actually predict experimental success.