23 February 2022 in IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Much of the recent work in remote sensing image captioning is influenced by natural image captioning.These methods tend to fix the defects of the model architecture to improve the previous work, but pay little attention to the differences between remote sensing images and natural images. By considering these differences, we propose a multi-scale multi-interaction remote sensing image captioning model. As in Figure~\ref{fig:intro} (a), the targets in remote sensing images have a wide range of scales; while the natural images are generally taken close-up, resulting in a similar scale for the foreground targets. Due to the difference in shooting methods, the model pre-trained on close-up natural images cannot capture multi-scale remote sensing targets well. To alleviate this problem, we propose a two-stage multi-scale structure for feature representation, where we first finetune the CNN backbone on remote sensing images for domain adaption, then we collect features from different stages as the multi-scale feature representation. Moreover, due to the shooting distance, the height information of the target in the remote sensing image is greatly weakened, thus some objects like low plants and grasses become difficult to identify, as in Figure~\ref{fig:intro} (b). Thus, we further propose a multi-interaction feature representation module, where information flow of the same and different layers could effectively interact. By calculating the similarity score among features, we fuse features with high similarity, and increase the distance between features of different categories, thereby enhancing the distinguishability. Results on RSICD, Sydney-Captions and UCM-Captions show a clear improvement over the compared methods.
30 December 2024 in Nature Communications
Thiamine and pyridoxine are essential B vitamins that serve as enzymatic cofactors in energy metabolism, protein and nucleic acid biosynthesis, and neurotransmitter production. In humans, thiamine transporters SLC19A2 and SLC19A3 primarily regulate cellular uptake of both vitamins. Genetic mutations in these transporters, which cause thiamine and pyridoxine deficiency, have been implicated in severe neurometabolic diseases. Additionally, various prescribed medicines, including metformin and fedratinib, manipulate thiamine transporters, complicating the therapeutic effect. Despite their physiological and pharmacological significance, the molecular underpinnings of substrate and drug recognition remain unknown. Here we present ten cryo-EM structures of human thiamine transporters SLC19A3 and SLC19A2 in outward- and inward-facing conformations, complexed with thiamine, pyridoxine, metformin, fedratinib, and amprolium. These structural insights, combined with functional characterizations, illuminate the translocation mechanism of diverse chemical entities, and enhance our understanding of drug-nutrient interactions mediated by thiamine transporters.
1 September 2021 in Knowledge-Based Systems
31 March 2026 in Nature Communications
ZYG11B is a substrate receptor of the Cullin2-RING E3 ligase (CRL2), mediating the Gly/N-degron pathway and contributing to diverse processes including cell cycle control, protein homeostasis, apoptosis, and innate immunity. While previous studies resolved the structure of its truncated ARM domain, how full-length ZYG11B coordinates substrate engagement and CRL2 ZYG11B assembly remains unclear. Here, we present cryo-EM structures of full-length human ZYG11B in complex with the EloB–EloC adaptor and a Gly/N-degron peptide, revealing a seahorse-like architecture with distinct interfaces for adaptor and substrate binding. Unexpectedly, ZYG11B adopts both monomeric and dimeric assemblies, with the dimer stabilizing two substrate-binding sites in opposite orientations. Functional assays demonstrate that interfaces mediating adaptor recruitment, substrate binding, and dimerization are essential for substrate degradation, suggesting a dynamic mechanism involving both assembly states. These findings provide a structural framework for understanding CRL2 ZYG11B -mediated ubiquitination and offer mechanistic insights that may inform the rational design of ZYG11B-based applications.