Jinan Institute of Quantum Technology

Abstract In recent years, cancer immunotherapy has gained substantial momentum, particularly in the field of vaccines and immunotherapeutics. Among these advancements, extracellular vesicles (EVs) have emerged as a promising cell‐free vaccine system due to their unique capacity to display diverse antigens and immune‐stimulatory molecules. Additionally, their excellent biocompatibility, ability to permeate across biological barriers, and targeting properties make EVs a promising platform for delivery of various therapeutics for cancer immunotherapy, such as RNAs, proteins, lipids, and chemotherapeutic agents. To date, EV‐based vaccines and immunotherapeutics have demonstrated remarkable efficacy in combating various types of cancers. Herein, we endeavor to give a comprehensive summary of EV‐based vaccines and immunotherapeutics for the treatment of cancers, delving into their advantages and challenges. Furthermore, emphasis was given on offering insight into the design rationale of each platform. Lastly, we recapitulate the recent preclinical and clinical application progress of EV‐based vaccines and immunotherapeutics, highlighting their pivotal role in cancer immunotherapy.

Macrocyclic peptides have gained attention as promising drug candidates due to their unique therapeutic properties. Advances in artificial intelligence have demonstrated the potential to facilitate the discovery and optimization of macrocyclic peptides. However, accurately predicting their biological activities in advance remains a significant challenge. In this study, we developed a multichannel predictive model that integrates molecular fingerprints, graph structural data, physicochemical characteristics, and ADMET properties. With the assistance of this model, we successfully identified macrocyclic peptides exhibiting potent inhibitory activity against neutrophil elastase and ADAM9. Validation was also performed on four independent peptide data sets. The results demonstrate a prediction accuracy of over 70% in unsupervised learning models and more than 90% with supervised learning models. This study provides a reliable multichannel machine learning model for predicting the bioactivity potential of macrocyclic peptides, demonstrating that the integration of a multichannel fusion strategy with machine learning can facilitate functional macrocyclic peptide screening.

In open‐world robotic manipulation tasks, language‐guided model‐free grasping has garnered increasing attention. However, existing approaches often overlook the geometric structure of target objects, which limits the effectiveness of subsequent tasks such as manipulation and placement. To address this limitation, a novel method called Language‐Guided Grasping via Primitive Fitting is proposed. This approach integrates language instructions with multimodal perception to enhance the semantic interpretability and downstream usability of the grasp through structured geometric modeling. Specifically, the user‐specified object using 2D images and depth data via multimodal understanding is first localized. Then, primitive fitting on the object's point cloud using basic geometric shapes (e.g., cuboids, ellipsoids, truncated cones) to extract approximate size and structural features is performed. Based on the geometric information, a grasp pose generation strategy guided by semantic geometry is defined, and modules for grasp feasibility filtering and task‐oriented optimization to select the optimal grasp pose are introduced. This method is validated in real‐world complex environments and achieved grasp success rates of 95% in structured and 90% in cluttered scenes. Geometric fitting enhances post‐grasp predictability and semantic consistency, enabling better generalization and planning.