Pub Date : 2026-02-28DOI: 10.1038/s41467-026-70171-2
Wen-Min Zhong, Wenbin Zhang, Yu-Xiang Zeng, JiYu Zhao, Ziqi Jia, Guanglong Ding, Su-Ting Han, Vellaisamy A. L. Roy, Ye Zhou
In the field of neuromorphic computing, time-series prediction poses a significant challenge to recurrent neural network architectures, often requiring task-specific customization that limits the development of general-purpose computing platforms. In this work, we implement a physical echo-state network (ESN) using ambipolar organic–inorganic heterostructure transistors to form its reservoir layer. Leveraging the ambipolar nature of the transistor, its variable-resistance region enables sparse matrix operations, while the saturation region provides tanh-like nonlinearity, making it well-suited for implementing both synaptic weighting and neuronal activation in an ESN. Additionally, its dynamic response naturally introduces temporal attributes. Thus, it can serve as a neuromorphic computing model for time-series tasks. Without the involvement of dynamic mechanisms, it is capable of performing image recognition, time-series prediction, and multimodal recognition tasks. When dynamic mechanisms are incorporated, the model achieves an accuracy of 96.98% on the MNIST handwritten digit dataset and 86.67% on the Fashion-MNIST dataset. This work offers a neuromorphic computing architecture, providing insights for tasks such as nonlinear mapping and time-series prediction.
{"title":"Physical echo state network based on the nonlinearity and dynamic response of ambipolar heterostructure transistors","authors":"Wen-Min Zhong, Wenbin Zhang, Yu-Xiang Zeng, JiYu Zhao, Ziqi Jia, Guanglong Ding, Su-Ting Han, Vellaisamy A. L. Roy, Ye Zhou","doi":"10.1038/s41467-026-70171-2","DOIUrl":"https://doi.org/10.1038/s41467-026-70171-2","url":null,"abstract":"In the field of neuromorphic computing, time-series prediction poses a significant challenge to recurrent neural network architectures, often requiring task-specific customization that limits the development of general-purpose computing platforms. In this work, we implement a physical echo-state network (ESN) using ambipolar organic–inorganic heterostructure transistors to form its reservoir layer. Leveraging the ambipolar nature of the transistor, its variable-resistance region enables sparse matrix operations, while the saturation region provides tanh-like nonlinearity, making it well-suited for implementing both synaptic weighting and neuronal activation in an ESN. Additionally, its dynamic response naturally introduces temporal attributes. Thus, it can serve as a neuromorphic computing model for time-series tasks. Without the involvement of dynamic mechanisms, it is capable of performing image recognition, time-series prediction, and multimodal recognition tasks. When dynamic mechanisms are incorporated, the model achieves an accuracy of 96.98% on the MNIST handwritten digit dataset and 86.67% on the Fashion-MNIST dataset. This work offers a neuromorphic computing architecture, providing insights for tasks such as nonlinear mapping and time-series prediction.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"350 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320105","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-70129-4
Shuai Zhang, Zhaolong Liu, Tongtong Jiang, Chen Wang, Jiahui Wang, Han Wang, Minqiang Fan, Li Yang, Yang Li, Liping Ding, Ying Yu, Xiaodong Hao, Shufang Ma, Bingshe Xu, Xiaolong Chen, Cong Ye, Xianfeng Chen, Paul K Chu, Shifeng Jin, Feng Ding, Xue-Feng Yu, Zhipei Sun, Jiahong Wang
Anisotropic materials with intrinsic one-dimensional architectures, where chains or tubes align along a crystallographic axis, exhibit direction-dependent optical responses and serve as ideal building blocks for polarization-sensitive optoelectronics. While progress exists in engineered compounds, discovering elemental crystals with naturally ordered one-dimensional building blocks exhibiting giant optical anisotropy remains challenging. Here, we report the synthesis of a direct-bandgap semiconducting one-dimensional phosphorus single crystal composed of unique wavy polygonal tubes. The monoclinic lattice structure is revealed by single-crystal X-ray diffraction and advanced transmission electron microscopy. The crystal exhibits giant birefringence in the visible and near-infrared regions, stemming from electron localization and anisotropic transitions of the phosphorus 3p orbital along the tube axis. The low-symmetry structure endows remarkable linear and nonlinear optical anisotropies, including orientation-dependent photoluminescence, Raman scattering, and second-harmonic generation. This study establishes a paradigm for designing giant optical anisotropies, opening avenues for on-chip polarization devices and nonlinear photonic circuits.
{"title":"Strong optical anisotropy in one-dimensional phosphorus wavy tubes.","authors":"Shuai Zhang, Zhaolong Liu, Tongtong Jiang, Chen Wang, Jiahui Wang, Han Wang, Minqiang Fan, Li Yang, Yang Li, Liping Ding, Ying Yu, Xiaodong Hao, Shufang Ma, Bingshe Xu, Xiaolong Chen, Cong Ye, Xianfeng Chen, Paul K Chu, Shifeng Jin, Feng Ding, Xue-Feng Yu, Zhipei Sun, Jiahong Wang","doi":"10.1038/s41467-026-70129-4","DOIUrl":"https://doi.org/10.1038/s41467-026-70129-4","url":null,"abstract":"<p><p>Anisotropic materials with intrinsic one-dimensional architectures, where chains or tubes align along a crystallographic axis, exhibit direction-dependent optical responses and serve as ideal building blocks for polarization-sensitive optoelectronics. While progress exists in engineered compounds, discovering elemental crystals with naturally ordered one-dimensional building blocks exhibiting giant optical anisotropy remains challenging. Here, we report the synthesis of a direct-bandgap semiconducting one-dimensional phosphorus single crystal composed of unique wavy polygonal tubes. The monoclinic lattice structure is revealed by single-crystal X-ray diffraction and advanced transmission electron microscopy. The crystal exhibits giant birefringence in the visible and near-infrared regions, stemming from electron localization and anisotropic transitions of the phosphorus 3p orbital along the tube axis. The low-symmetry structure endows remarkable linear and nonlinear optical anisotropies, including orientation-dependent photoluminescence, Raman scattering, and second-harmonic generation. This study establishes a paradigm for designing giant optical anisotropies, opening avenues for on-chip polarization devices and nonlinear photonic circuits.</p>","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":" ","pages":""},"PeriodicalIF":15.7,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147317720","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-69882-3
Digant Nayak, Lijia Jia, Priscila dos Santos Bury, Eliza A. Ruben, Ankita Shukla, Anindita Nayak, Caleb M. Stratton, Pirouz Ebadi, Hee Cho, Anna A. Tumanova, Joyce T. Varughese, Lingmin Yuan, Fei Gao, Kristin E. Cano, Christopher Davies, Patrick Sung, Michaela U. Gack, Elizabeth V. Wasmuth, Shaun K. Olsen
UBA1 and UBA6 define parallel ubiquitin (Ub) activation systems that perform non-overlapping roles in Ub and ubiquitin-like protein (Ubl) signaling. Whereas UBA1 supports the canonical Ub pathway, UBA6 also activates the Ubl FAT10, linking Ub signaling to immune-regulated proteostasis. In addition to selective Ub/Ubl activation, UBA1 and UBA6 engage distinct sets of E2s, yet how these enzymes achieve selective E2 engagement has remained unclear. Using chemical trapping and high-resolution cryo-EM, we determine four structures of UBA6–E2 complexes representing the thioester-transfer step with either FAT10 or Ub, revealing how this E1 distinguishes its cognate partners. UBA6 achieves E2 specificity through coordinated contributions of the UFD and SCCH domains, a dual-domain mechanism that contrasts with the UFD-dominated selectivity of UBA1. The structures further show that an existing inositol hexakisphosphate (InsP₆)–binding site, unique to UBA6, stabilizes an expanded SCCH cleft that pre-organizes the enzyme for selective engagement of UBA6-specific E2s. These findings define principles for E1–E2 recognition and identify InsP₆ as a cofactor shaping specificity within the Ub-like conjugation network.
{"title":"Cryo-EM structures of UBA6 reveal mechanisms of E1–E2 specificity and dual FAT10/ubiquitin thioester transfer","authors":"Digant Nayak, Lijia Jia, Priscila dos Santos Bury, Eliza A. Ruben, Ankita Shukla, Anindita Nayak, Caleb M. Stratton, Pirouz Ebadi, Hee Cho, Anna A. Tumanova, Joyce T. Varughese, Lingmin Yuan, Fei Gao, Kristin E. Cano, Christopher Davies, Patrick Sung, Michaela U. Gack, Elizabeth V. Wasmuth, Shaun K. Olsen","doi":"10.1038/s41467-026-69882-3","DOIUrl":"https://doi.org/10.1038/s41467-026-69882-3","url":null,"abstract":"UBA1 and UBA6 define parallel ubiquitin (Ub) activation systems that perform non-overlapping roles in Ub and ubiquitin-like protein (Ubl) signaling. Whereas UBA1 supports the canonical Ub pathway, UBA6 also activates the Ubl FAT10, linking Ub signaling to immune-regulated proteostasis. In addition to selective Ub/Ubl activation, UBA1 and UBA6 engage distinct sets of E2s, yet how these enzymes achieve selective E2 engagement has remained unclear. Using chemical trapping and high-resolution cryo-EM, we determine four structures of UBA6–E2 complexes representing the thioester-transfer step with either FAT10 or Ub, revealing how this E1 distinguishes its cognate partners. UBA6 achieves E2 specificity through coordinated contributions of the UFD and SCCH domains, a dual-domain mechanism that contrasts with the UFD-dominated selectivity of UBA1. The structures further show that an existing inositol hexakisphosphate (InsP₆)–binding site, unique to UBA6, stabilizes an expanded SCCH cleft that pre-organizes the enzyme for selective engagement of UBA6-specific E2s. These findings define principles for E1–E2 recognition and identify InsP₆ as a cofactor shaping specificity within the Ub-like conjugation network.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"83 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320049","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-70106-x
Qilong Bian, Ying Zhao, Chunhua Zhang, Yang Zhang, Yang Zhang, Zebing Zeng, Ben Zhong Tang, Sheng Xie
Topology plays an important role in polymeric materials. Herein, we present an iterative, modular approach for creating tetra-arylsubstituted alkene (TAA)-based dynamic conjugated oligomers with diverse topologies, using boronate-protected Suzuki-Miyaura coupling chemistry. The TAA building blocks involving spontaneous alkene isomerization are found to induce conformational dynamics in the conjugated backbones, exhibiting steric-controlled transitions. These transitions occur from a twisted backbone rich in cis-alkenes in the linear PL9 oligomer, to a stretched backbone with a trans-alkene center and multiple cis-alkene end in the three-armed planar PY12 oligomer and the four-armed 3D PX16 oligomer. Consequently, these topological oligomers exhibit distinct photoluminescence and photochemical properties depending on their physical state. Experimental characterization and molecular dynamics simulations (MD) reveal a topology-dependent adaptive self-assembly of helices: linear PL9 forms long flexible helical fibers with a pitch of 28 nm; planar Y-type PY12 oligomers often occur in neural-like networks, connected by nanofibers and cell-like central aggregates; and stereo X-type PX16 adopts short helical rod-like morphology with a mesoscopic pitch of 86 nm in crystalline phases. This work may inspire concepts and the practical construction of helical and neural-like fiber materials by altering unit topology in dynamic conjugated oligomers.
{"title":"Topology-controlled dynamic conjugated oligomers from tetra-arylsubstituted alkene building blocks","authors":"Qilong Bian, Ying Zhao, Chunhua Zhang, Yang Zhang, Yang Zhang, Zebing Zeng, Ben Zhong Tang, Sheng Xie","doi":"10.1038/s41467-026-70106-x","DOIUrl":"https://doi.org/10.1038/s41467-026-70106-x","url":null,"abstract":"Topology plays an important role in polymeric materials. Herein, we present an iterative, modular approach for creating tetra-arylsubstituted alkene (TAA)-based dynamic conjugated oligomers with diverse topologies, using boronate-protected Suzuki-Miyaura coupling chemistry. The TAA building blocks involving spontaneous alkene isomerization are found to induce conformational dynamics in the conjugated backbones, exhibiting steric-controlled transitions. These transitions occur from a twisted backbone rich in cis-alkenes in the linear PL9 oligomer, to a stretched backbone with a trans-alkene center and multiple cis-alkene end in the three-armed planar PY12 oligomer and the four-armed 3D PX16 oligomer. Consequently, these topological oligomers exhibit distinct photoluminescence and photochemical properties depending on their physical state. Experimental characterization and molecular dynamics simulations (MD) reveal a topology-dependent adaptive self-assembly of helices: linear PL9 forms long flexible helical fibers with a pitch of 28 nm; planar Y-type PY12 oligomers often occur in neural-like networks, connected by nanofibers and cell-like central aggregates; and stereo X-type PX16 adopts short helical rod-like morphology with a mesoscopic pitch of 86 nm in crystalline phases. This work may inspire concepts and the practical construction of helical and neural-like fiber materials by altering unit topology in dynamic conjugated oligomers.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"15 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320096","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
It is very difficult to prepare vertical nanodiamond (VND) dominated sheets compared with vertical graphene (VG) sheets. Here, we describe the preparation of VNDs by phase transformation from Ta-loaded VGs via argon/oxygen plasma treatment with varied oxygen percentage. As the oxygen percentage increases to 5%, the height of VGs decreases, accompanied by high capacitance and low Hall mobility. When the oxygen percentage increases to 10%, the VGs transform to nanocrystalline diamond (NCD) grains with trans-polyacetylene (TPA) contents in grain boundaries increasing significantly. These VNDs exhibit both high capacitance (1452 μF cm−2) and n-type Hall mobility (846 cm2 V−1s−1). With the oxygen percentage increasing to 20%, NCD grains grow larger and the TPA content reduces, while the capacitance decreases considerably, but Hall mobility remaining high, suggesting that the performance improvement results from the synergistic effect of NCDs and TPA. These exhibit significant applications of VNDs in energy storage, high performance sensors, high-frequency and high-power electronic devices.
{"title":"Vertical nanodiamond dominated sheets possessing both high capacitance and high n-type Hall mobility","authors":"Yuemin Gong, Zhiqiang Zhang, Meiyan Jiang, Chengke Chen, Shaohua Lu, Xiaojun Hu","doi":"10.1038/s41467-026-70089-9","DOIUrl":"https://doi.org/10.1038/s41467-026-70089-9","url":null,"abstract":"It is very difficult to prepare vertical nanodiamond (VND) dominated sheets compared with vertical graphene (VG) sheets. Here, we describe the preparation of VNDs by phase transformation from Ta-loaded VGs via argon/oxygen plasma treatment with varied oxygen percentage. As the oxygen percentage increases to 5%, the height of VGs decreases, accompanied by high capacitance and low Hall mobility. When the oxygen percentage increases to 10%, the VGs transform to nanocrystalline diamond (NCD) grains with trans-polyacetylene (TPA) contents in grain boundaries increasing significantly. These VNDs exhibit both high capacitance (1452 μF cm−2) and n-type Hall mobility (846 cm2 V−1s−1). With the oxygen percentage increasing to 20%, NCD grains grow larger and the TPA content reduces, while the capacitance decreases considerably, but Hall mobility remaining high, suggesting that the performance improvement results from the synergistic effect of NCDs and TPA. These exhibit significant applications of VNDs in energy storage, high performance sensors, high-frequency and high-power electronic devices.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"129 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320130","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-69816-z
Zhiwen Jiang, Jason Stein, Tengfei Li, Ethan Fang, Yun Li, Patrick Sullivan, Hongtu Zhu
Imaging genetics links genetic variations to brain structures and functions, but the computational challenges posed by high-dimensional imaging and genetic data are significant. In voxel-level genome-wide association studies, we introduce a Representation learning-based Voxel-level Genetic Analysis (RVGA) framework that reduces computational time and storage burden by over 200 times. RVGA enhances statistical power by denoising images and shares minimal datasets of summary statistics for associations across the whole genome of the entire image for secondary analyses. Additionally, it introduces a unified estimator for voxel heritability, genetic correlations between voxels, and cross-trait genetic correlations between voxels and non-imaging phenotypes. Applying RVGA to hippocampus shape and white matter microstructure in the UK Biobank (n = 53,454) reveals 39 and 275 novel loci, respectively. We identify heterogeneity in heritability within images and subregions that share genetic bases with 14 brain-related phenotypes, such as the genetic correlation between the hippocampus and educational attainment, and between the anterior corona radiata and schizophrenia. RVGA replicates known genetic associations and uncovers new discoveries.
{"title":"Computation and resource efficient genome-wide association analysis for large-scale imaging studies","authors":"Zhiwen Jiang, Jason Stein, Tengfei Li, Ethan Fang, Yun Li, Patrick Sullivan, Hongtu Zhu","doi":"10.1038/s41467-026-69816-z","DOIUrl":"https://doi.org/10.1038/s41467-026-69816-z","url":null,"abstract":"Imaging genetics links genetic variations to brain structures and functions, but the computational challenges posed by high-dimensional imaging and genetic data are significant. In voxel-level genome-wide association studies, we introduce a Representation learning-based Voxel-level Genetic Analysis (RVGA) framework that reduces computational time and storage burden by over 200 times. RVGA enhances statistical power by denoising images and shares minimal datasets of summary statistics for associations across the whole genome of the entire image for secondary analyses. Additionally, it introduces a unified estimator for voxel heritability, genetic correlations between voxels, and cross-trait genetic correlations between voxels and non-imaging phenotypes. Applying RVGA to hippocampus shape and white matter microstructure in the UK Biobank (n = 53,454) reveals 39 and 275 novel loci, respectively. We identify heterogeneity in heritability within images and subregions that share genetic bases with 14 brain-related phenotypes, such as the genetic correlation between the hippocampus and educational attainment, and between the anterior corona radiata and schizophrenia. RVGA replicates known genetic associations and uncovers new discoveries.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"125 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-69730-4
Yufei Li, Kang Li, Jiazhi Lu, Mustafa Bulut, Huanteng Hou, Qamar U. Zaman, Ran Zhang, Zhuang Yang, Chenkun Yang, Chuansong Zhan, Guan Wang, Tong Chen, Xianqing Liu, Qiao Zhao, Shuangqian Shen, Alisdair R. Fernie, Jie Luo
While enhancing crop yield remains a critical breeding objective, such efforts often compromise nutritional quality and stress tolerance. The growing global population and increasing environmental stresses make combining stable crop yields with improved nutritional quality a vital objective for sustainable food security. Thiamine pyrophosphate (TPP), the active form of vitamin B1, regulates core metabolism to support both human health and plant stress tolerances. Here, we show that editing the rice TPP riboswitch elevates multiple micronutrients, simultaneously increasing grain yield, cold tolerance, and blast resistance. Mechanistically, these enhancements link to promoted photosynthesis, improved nitrogen use efficiency, and system-wide transcriptional-metabolic reprogramming. Editing homologs in tomato produces similar outcomes, supporting the generalizability of this approach. Thus, modulating TPP levels offers a viable strategy to synergistically boost crop productivity, nutritional quality, and stress tolerance, which are essential for achieving sustainable food security without trade-offs.
{"title":"Modulating TPP riboswitch activity simultaneously enhances crop yield, nutritional quality and stress tolerance","authors":"Yufei Li, Kang Li, Jiazhi Lu, Mustafa Bulut, Huanteng Hou, Qamar U. Zaman, Ran Zhang, Zhuang Yang, Chenkun Yang, Chuansong Zhan, Guan Wang, Tong Chen, Xianqing Liu, Qiao Zhao, Shuangqian Shen, Alisdair R. Fernie, Jie Luo","doi":"10.1038/s41467-026-69730-4","DOIUrl":"https://doi.org/10.1038/s41467-026-69730-4","url":null,"abstract":"While enhancing crop yield remains a critical breeding objective, such efforts often compromise nutritional quality and stress tolerance. The growing global population and increasing environmental stresses make combining stable crop yields with improved nutritional quality a vital objective for sustainable food security. Thiamine pyrophosphate (TPP), the active form of vitamin B1, regulates core metabolism to support both human health and plant stress tolerances. Here, we show that editing the rice TPP riboswitch elevates multiple micronutrients, simultaneously increasing grain yield, cold tolerance, and blast resistance. Mechanistically, these enhancements link to promoted photosynthesis, improved nitrogen use efficiency, and system-wide transcriptional-metabolic reprogramming. Editing homologs in tomato produces similar outcomes, supporting the generalizability of this approach. Thus, modulating TPP levels offers a viable strategy to synergistically boost crop productivity, nutritional quality, and stress tolerance, which are essential for achieving sustainable food security without trade-offs.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"99 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-69845-8
Wenxin Zhang, Zhi Li, Huajian Gao, Julia R Greer
3D printing methods for small-scale metals enable a unique 10-100 nm dimensional niche where functional feature sizes, critical microstructural detail and atomic-level defects converge, challenging conventional hierarchical relationships and carrying significant nanomechanical implications. We introduce a metal nano-printing system combining two-photon lithography, hydrogel infusion-based additive manufacturing and in situ mechanical experiments on 3D nano-architected Ni, achieving ~100 nm critical dimensions, ~10 nm surface roughness, and a broad range of geometries (periodic vs. non-periodic; beam-based vs. shell-based) with superior specific strengths of ~100 MPa·g - 1·cm3 enabled by an unambiguous smaller is stronger size effect. Experiments identify concentrated-porosity regions as primary deformation-initiation sources and quantify their distribution as input for physics-informed, multiscale finite-element simulations that accurately predict size-dependent mechanical properties governed by nanoporosity-driven deformation. This work integrates experimental and computational approaches for the fabrication, characterization, and evaluation of nano- and micro-architected metals for nanotechnology and nanoscale manufacturing systems.
{"title":"Nanoporosity-driven deformation of additively manufactured nano-architected metals.","authors":"Wenxin Zhang, Zhi Li, Huajian Gao, Julia R Greer","doi":"10.1038/s41467-026-69845-8","DOIUrl":"https://doi.org/10.1038/s41467-026-69845-8","url":null,"abstract":"<p><p>3D printing methods for small-scale metals enable a unique 10-100 nm dimensional niche where functional feature sizes, critical microstructural detail and atomic-level defects converge, challenging conventional hierarchical relationships and carrying significant nanomechanical implications. We introduce a metal nano-printing system combining two-photon lithography, hydrogel infusion-based additive manufacturing and in situ mechanical experiments on 3D nano-architected Ni, achieving ~100 nm critical dimensions, ~10 nm surface roughness, and a broad range of geometries (periodic vs. non-periodic; beam-based vs. shell-based) with superior specific strengths of ~100 MPa·g - 1·cm<sup>3</sup> enabled by an unambiguous smaller is stronger size effect. Experiments identify concentrated-porosity regions as primary deformation-initiation sources and quantify their distribution as input for physics-informed, multiscale finite-element simulations that accurately predict size-dependent mechanical properties governed by nanoporosity-driven deformation. This work integrates experimental and computational approaches for the fabrication, characterization, and evaluation of nano- and micro-architected metals for nanotechnology and nanoscale manufacturing systems.</p>","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":" ","pages":""},"PeriodicalIF":15.7,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147317901","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-70077-z
Yukun Zhou, Zheyuan Wang, Yilan Wu, Ariel Yuhan Ong, Siegfried K. Wagner, Eden Ruffell, Mark A. Chia, Zhouyu Guan, Lie Ju, Justin Engelmann, David A. Merle, Tingyao Li, Jia Shu, Paul Nderitu, Ke Zou, Jocelyn Hui Lin Goh, Qingshan Hou, Xiaoxuan Liu, Yaxing Wang, Yih Chung Tham, Andre Altmann, Carol Y. Cheung, Daniel C. Alexander, Eric J. Topol, Alastair K. Denniston, Tien Yin Wong, Bin Sheng, Pearse A. Keane
Medical foundation models, pre-trained on large-scale unlabelled data, show strong performance and data efficiency when adapted to various clinically relevant applications. However, how pre-training data shape the generalisability and fairness of these models remains unexplored. Here we address this using two cohorts from Moorfields Eye Hospital (UK) and the Shanghai Diabetes Prevention Program (China), each containing 904,170 fundus photographs for model pre-training. Using identical pipelines, we train parallel foundation models using individual cohort and evaluate them on downstream tasks with publicly available datasets and held-out data from each site. The parallel models show competitive performance to data that differ substantially from their pre-training data. Nevertheless, we observe fairness gaps over age subgroups, whereas sex and ethnicity show minimal impact. These results demonstrate the good generalisability of retinal foundation models and indicate that pre-training demographic attributes shape fairness differently, highlighting the importance of domain-specific, fine-grained data curation for efficient foundation model development.
{"title":"Understanding pre-training data effects in retinal foundation models using two large fundus cohorts","authors":"Yukun Zhou, Zheyuan Wang, Yilan Wu, Ariel Yuhan Ong, Siegfried K. Wagner, Eden Ruffell, Mark A. Chia, Zhouyu Guan, Lie Ju, Justin Engelmann, David A. Merle, Tingyao Li, Jia Shu, Paul Nderitu, Ke Zou, Jocelyn Hui Lin Goh, Qingshan Hou, Xiaoxuan Liu, Yaxing Wang, Yih Chung Tham, Andre Altmann, Carol Y. Cheung, Daniel C. Alexander, Eric J. Topol, Alastair K. Denniston, Tien Yin Wong, Bin Sheng, Pearse A. Keane","doi":"10.1038/s41467-026-70077-z","DOIUrl":"https://doi.org/10.1038/s41467-026-70077-z","url":null,"abstract":"Medical foundation models, pre-trained on large-scale unlabelled data, show strong performance and data efficiency when adapted to various clinically relevant applications. However, how pre-training data shape the generalisability and fairness of these models remains unexplored. Here we address this using two cohorts from Moorfields Eye Hospital (UK) and the Shanghai Diabetes Prevention Program (China), each containing 904,170 fundus photographs for model pre-training. Using identical pipelines, we train parallel foundation models using individual cohort and evaluate them on downstream tasks with publicly available datasets and held-out data from each site. The parallel models show competitive performance to data that differ substantially from their pre-training data. Nevertheless, we observe fairness gaps over age subgroups, whereas sex and ethnicity show minimal impact. These results demonstrate the good generalisability of retinal foundation models and indicate that pre-training demographic attributes shape fairness differently, highlighting the importance of domain-specific, fine-grained data curation for efficient foundation model development.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"22 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2026-02-28DOI: 10.1038/s41467-026-70069-z
Runzhi Xie, Zekun Wang, Jianrui Liu, Han Xu, Yafei Xu, Jiadong Lin, Yuxuan Hu, Lin Gao
Tissues are organized through the assembly of diverse cell types into multicellular structures that exhibit hierarchical spatial organization. We present HRCHY-CytoCommunity, a graph neural network framework for identifying multi-level tissue structures directly from cell-type annotated spatial maps. It integrates differentiable graph pooling, adaptive edge pruning, and consistency and balance regularization in an end-to-end model, simultaneously inferring robust structures across multiple scales while preserving complete cellular coverage and fully nested relationships. The framework also supports cross-sample hierarchy alignment via cell-type enrichment-based clustering. Benchmarking on diverse spatial omics datasets, HRCHY-CytoCommunity outperforms existing hierarchical and non-hierarchical methods in identifying both coarse-grained tissue compartments and fine-grained cellular neighborhoods. Applied to a breast cancer cohort with clinical outcomes, the framework enables hierarchical prognostic stratification of patients and reveals survival-associated spatial patterns. HRCHY-CytoCommunity represents a general and scalable tool for deciphering tissue organization from single cells to multicellular modules, and ultimately to intact tissues and organs.
{"title":"HRCHY-CytoCommunity identifies hierarchical tissue organization in cell-type spatial maps","authors":"Runzhi Xie, Zekun Wang, Jianrui Liu, Han Xu, Yafei Xu, Jiadong Lin, Yuxuan Hu, Lin Gao","doi":"10.1038/s41467-026-70069-z","DOIUrl":"https://doi.org/10.1038/s41467-026-70069-z","url":null,"abstract":"Tissues are organized through the assembly of diverse cell types into multicellular structures that exhibit hierarchical spatial organization. We present HRCHY-CytoCommunity, a graph neural network framework for identifying multi-level tissue structures directly from cell-type annotated spatial maps. It integrates differentiable graph pooling, adaptive edge pruning, and consistency and balance regularization in an end-to-end model, simultaneously inferring robust structures across multiple scales while preserving complete cellular coverage and fully nested relationships. The framework also supports cross-sample hierarchy alignment via cell-type enrichment-based clustering. Benchmarking on diverse spatial omics datasets, HRCHY-CytoCommunity outperforms existing hierarchical and non-hierarchical methods in identifying both coarse-grained tissue compartments and fine-grained cellular neighborhoods. Applied to a breast cancer cohort with clinical outcomes, the framework enables hierarchical prognostic stratification of patients and reveals survival-associated spatial patterns. HRCHY-CytoCommunity represents a general and scalable tool for deciphering tissue organization from single cells to multicellular modules, and ultimately to intact tissues and organs.","PeriodicalId":19066,"journal":{"name":"Nature Communications","volume":"23 1","pages":""},"PeriodicalIF":16.6,"publicationDate":"2026-02-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147320150","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：