The field of molecular design and generation is rapidly advancing, with a focus on developing innovative methods for improving the efficiency and accuracy of molecular geometry generation, protein design, and mRNA sequence generation. Recent developments have centered around the use of deep learning frameworks, such as neural pipeline approaches and generative models, to enhance the design and generation of molecular structures. These approaches have shown significant promise in addressing challenges such as synthesis costs, sequencing errors, and biological constraints. Notably, the integration of biologically informed constraints with deep learning has enabled the development of more robust and accurate methods for molecular design and generation. Furthermore, the use of techniques such as score distillation and curriculum learning has improved the efficiency and effectiveness of these methods. Overall, the field is moving towards the development of more scalable, biologically valid, and accurate methods for molecular design and generation. Noteworthy papers include: NEURODNAAI, which achieves superior accuracy in DNA storage by integrating biologically informed constraints with deep learning. Distilled Protein Backbone Generation, which reduces sampling time by over 20-fold while maintaining comparable performance to its pretrained teacher model. Curriculum-Augmented GFlowNets For mRNA Sequence Generation, which improves Pareto performance and biological plausibility in mRNA design tasks. Predictive Feature Caching for Training-free Acceleration of Molecular Geometry Generation, which achieves a twofold reduction in wall-clock inference time at matched sample quality. Flow-Matching Based Refiner for Molecular Conformer Generation, which improves sample quality with fewer total denoising steps while preserving diversity.