The field of protein design and engineering is rapidly advancing, with a focus on developing innovative methods for designing and optimizing protein structures and functions. Recent developments have highlighted the potential of reinforcement learning, graph neural networks, and deep learning frameworks for enhancing protein thermostability and designing enzymes with desired thermal properties. The use of 3D pharmacophores and generative models has also shown promise in structure-based drug design. Furthermore, the development of benchmarks and frameworks for evaluating and comparing different methods has enabled more comprehensive investigations and fair comparisons of various approaches. Notable papers in this area include: ThermoRL, which presents a framework for designing mutations with enhanced thermostability using reinforcement learning and graph neural networks. The Segment Transformer, which achieves state-of-the-art performance in predicting enzyme temperature stability. MEVO, which efficiently generates high-affinity binders for various protein targets using a diffusion model and pocket-aware evolutionary strategy. Protein-SE(3), which provides a comprehensive benchmark for SE(3)-based generative models for protein structure design. Adam-PnP, which guides a pre-trained protein diffusion model using gradients from multiple experimental sources. DynamicMPNN, which generates sequences compatible with multiple conformations through joint learning across conformational ensembles.