The field of protein structure prediction and design is rapidly advancing, with a focus on developing innovative methods that integrate machine learning, biophysical models, and experimental techniques. Recent developments have highlighted the potential of hybrid approaches that combine the strengths of different methods to improve the accuracy and efficiency of protein structure prediction and design. One notable trend is the use of deep learning models to reconstruct 3D protein structures from multi-view data, such as AFM images, and to generate novel protein sequences with desired properties. Another area of research is the development of simpler, more general-purpose architectures for protein folding models, which have shown competitive performance compared to state-of-the-art baselines. Additionally, multimodal machine learning frameworks are being explored for therapeutic antibody reformatting, with surprising results showing that simple, domain-tailored representations can outperform large pretrained protein language models. Noteworthy papers include: ProFusion, which presents a hybrid framework for 3D reconstruction of protein complex structures from multi-view AFM images. Monte Carlo Tree Diffusion with Multiple Experts, which proposes a novel framework for protein design that integrates masked diffusion models with tree search. SimpleFold, which introduces a flow-matching based protein folding model that solely uses general-purpose transformer blocks. Improved Therapeutic Antibody Reformatting through Multimodal Machine Learning, which develops a machine learning framework to predict reformatting success and prioritization of promising candidates.