If you cannot view this email properly, please click here.

Welcome to InnoSci, the research newsletter of the School of Science at the Hong Kong University of Science and Technology. Through this newsletter, we are excited to share with the academic community the recent research discoveries and achievements of our faculty members, showcasing their novel ideas and unwavering commitment to scientific advancement. This newsletter will serve as a platform for exchanging insights and fostering collaborations with researchers around the globe in the pursuit of scientific excellence and innovation. Breakthrough in Light Manipulation Using Gyromagnetic Zero-Index Metamaterials An      HKUST- led      research      team,      co-led      by Prof. CHAN Che-Ting and Dr. ZHANG Ruoyang from the Department of Physics, has adopted gyromagnetic double-zero-index metamaterials (GDZIMs) – a new optical extreme-parameter material – and developed a groundbreaking method to control light using GDZIMs. Through constructing a magnetic photonic crystal and tuning the parameters to the critical phase transition point, the researchers realized this metamaterial for the first time. Their investigation revealed that the generation of the light vortex stems from GDZIM’s intrinsic topological properties, thereby ensuring exceptional stability regardless of the material’s size or surrounding environment. This breakthrough could lead to significant improvements in optical technologies, such as faster and more secure communication systems. Nature Volume 641, pages1142–1148 (2025) DOI: https://doi.org/10.1038/s41586-025-08948-6 Unveiling Boracycles: A New Approach to Synthesizing Strained Boracycles A research team led by Prof. QUAN Yangjian and Prof. LIN Zhenyang from the Department of Chemistry at HKUST, in collaboration with Prof. LYU Hairong from the Chinese University of Hong Kong (CUHK), has made a breakthrough in developing an efficient synthetic approach to four-membered boracycles, which are largely unexplored in current studies. In this recent research, a balance between the stability and reactivity of four-membered boracycles has been achieved for the first time, enabling the facile synthesis of previously inaccessible boracycles. These new boracycles are expected to have valuable applications in boron pharmaceuticals and molecular functional materials. Nature Chemistry Volume 17, pages663–671 (2025) DOI: https://doi.org/10.1038/s41557-025-01807-x Advancing Plant Gene Regulation with GHR Motif Discovery A research team led by Prof. NGUYEN Tuan Anh from the Division of Life Science at HKUST has recently achieved significant progress in understanding microRNA (miRNA) biogenesis. They developed a groundbreaking massively parallel dicing assay to investigate miRNA-producing enzymes in humans (DICER) and plants (DCL1). In their recent study, the team identified the conserved GHR motif as a critical factor determining DCL1's cleavage specificity in plants. This innovative assay and motif identification approach can be extended to the study of other DICER-like enzymes, advancing our knowledge of RNA-based gene regulation. Nature Plants Volume 11, pages1528–1543 (2025) DOI: https://doi.org/10.1038/s41477-025-02067-w Novel AI-powered Tool for Accurate Prediction of Coastal Oceans’ Health A research team led by Prof. GAN Jianping from the Department of Ocean Science and Prof. YANG Can from the Department of Mathematics at HKUST, has developed a novel AI-powered tool named STIMP for diagnosing coastal ocean productivity and ecosystem health. STIMP introduces a novel paradigm that imputes missing data and then predicts Chlorophyll-a (Chl-a) concentrations across large spatiotemporal scales. In tests across four representative global coastal regions, STIMP significantly outperformed existing geoscience tools, reducing the mean absolute error (MAE) for imputation by up to 81.39% and for prediction by 58.99%. Accurate Chl-a prediction aids in early detection of harmful algal blooms, ecosystem protection, and provides data-driven insights for evidence-based policy-making. Nature Communications Volume 16, 7656 (2025) DOI: https://doi.org/10.1038/s41467-025-62901-9 Uncovering Flow State Transitions in Active Fluids via Viscoelastic Modeling Active fluids encompass a wide range of non-equilibrium fluids, where the self-propulsion or rotation of their units can give rise to large-scale spontaneous flows. Despite this diversity, they are commonly viscoelastic. Prof. QIAN Tiezheng from the Department of Mathematics at HKUST and the research team led by Prof. ZHANG Rui from the Department of Physics have developed a hydrodynamic model of isotropic active liquids that incorporates an active stress term into a general viscoelastic liquid model. This model allows for the study of spontaneous flow states and their transitions in various geometrics, including channels, annuals and disks, uncover their connections and highlights the universality of these dynamic active-flow patterns, revealing the important role activity and Weissenberg number play. Journal of Fluid Mechanics Volume 7, R7 DOI: https://doi.org/10.1017/jfm.2025.177 Novel Vesicle-Based Method Advancing Membrane Protein Research Scientists have used detergents for decades to extract membrane proteins for structural studies, but this method has limitations, including resource-intensive screening and the lack of native lipids, which can impede research on lipid-mediated regulation. To   address   these   challenges,   a   research   team   led   by Prof. DANG Shangyu from the Division of Life Science at HKUST has developed a novel vesicle-based method that preserves the native lipid environment of membrane proteins, which can advance structural and functional studies. After four years of systematic investigation, Prof. Dang’s team bypassed the need for purification with detergents by directly generating vesicles containing the target protein from cell membranes. This approach produced samples suitable for cryo-EM imaging and structural studies. The team established a comprehensive workflow for the preparation, purification, and quality control of vesicle samples, making this method applicable to various membrane systems. PNAS Volume 122 (36) e2423407122 DOI: https://doi.org/10.1073/pnas.2423407122 Discovery on a “Super” Photoreductant to Leap Forward in Photocatalysis A research team led by Prof. LU Haipeng from the Department of Chemistry at HKUST have recently made a significant breakthrough in photocatalysis by unveiling a “super” photoreductant, marking a major advancement in organic synthesis. The team has developed an innovative photocatalytic system that utilizes visible-light-absorbing QDs, and introduced an efficient hot-electron generation mechanism facilitated by the two-photon spin-exchange Auger process in Mn²⁺-doped CdS/ZnS QDs. The study underscores the unprecedented potential of quantum-confined semiconductors to facilitate challenging organic transformations that were previously unattainable with conventional molecular photocatalysts. Nature Communications Volume 16 (5280) (2025) DOI: https://doi.org/10.1038/s41467-025-60659-8 Two niobium superconducting electrodes are connected by a magnetic topological superconductor MnBi2Te4. The maximum dissipationless current flowing from left to right is different from when the current flows in opposite direction, forming a superconducting diode device. Development of a Nonvolatile Josephson Diode with Programmable Polarity A research team led by Prof. LAW Kam Tuen from the Department of Physics at HKUST has made significant advancements in condensed matter physics by hybridizing superconductivity with topology and magnetism. They realized a nonvolatile Josephson diode device with programmable polarity using MnBi2Te4. The study establishes a system as a promising avenue for investigating topological superconductivity and their applications in superconducting electronic devices. Science Advances Volume 11, Issue 20 DOI: https://doi.org/10.1126/sciadv.ads8730 Mapping the Molecular Landscape of Grey Mouse Lemurs Prof. Angela Ruohao WU's team (Division of Life Science and Department of Chemical and Biological Engineering) and Prof. YANG Can's team (Department of Mathematics) at HKUST collaborated with other researchers around the world to create a "cell atlas" that maps over 226,000 single cells from 27 organs of the grey mouse lemur - a non-human primate native to the rainforests of Madagascar. They have created a comprehensive cellular and molecular atlas of mouse lemurs, the smallest and fastest reproducing primates, using advanced RNA sequencing techniques. This atlas identifies over 750 cell types and thousands of previously unknown genes, providing valuable insights into primate evolution, immune systems, and diseases, including conditions similar to human cancers. The findings establish mouse lemurs as an important model for studying primate biology and genetics, reshaping the way for future research into human biology and disease. Nature Volume 644, pages173–184 (2025) DOI: https://doi.org/10.1038/s41586-025-09113-9 Nature Volume 644, pages185–196 (2025) DOI: https://doi.org/10.1038/s41586-025-09114-8 Innovation in the Synthesis of Chiral Bridged Polycyclic Compounds An  HKUST  research  team  led  by  Prof. SUN  Jianwei  and Prof. LIN Zhenyang from the Department of Chemistry has developed a catalytic enantioselective type II [5 + 2] cycloaddition method to address the challenges of synthesizing chiral bridged polycyclic structures, particularly those with a bridged seven-membered subunit. This innovative approach utilizes 3-oxidopyrylium ylides to create the desired complex shapes, paving the way for more applications in the rapid synthesis and diversification of other valuable complex molecules, including important natural products and drug molecules. Nature Synthesis (2025) DOI: https://doi.org/10.1038/s44160-025-00803-w Innovative Technology to Identify cfRNA Modification Biomarkers for Early Cancer Detection A research team led by Prof. ZHANG Li-Sheng from the Division of Life Science and the Department of Chemistry at HKUST, has developed a significant cancer detection method called LIME-seq (Low-Input Multiple Methylation Sequencing). This innovative approach focuses on analyzing circulating free RNA (cfRNA) in the blood, which has the potential to improve early cancer diagnosis, particularly for challenging cases like colorectal cancer. LIME-seq has three key features: it requires less than 2 ng of RNA for sequencing, detects a wide range of RNA modifications including m1A and m3C, and utilizes the “read-through” capability of HIV reverse transcriptase to convert RNA modifications into identifiable mutation signals for precise localization and quantification. Nature Biotechnology (2025) DOI: https://doi.org/10.1038/s41587-025-02731-8 Automatic differentiation in training neural networks for solving differential equations Prof. XIANG Yang from the Department of Mathematics at HKUST has a recent breakthrough on developing accurate and efficient neural network methods for solving partial differential equations (PDEs). Neural network-based approaches have recently shown significant promise in solving PDEs in science and engineering, especially in scenarios featuring complex domains, incorporation of empirical data, and high dimensional problems. The research team led by Prof. Xiang has successfully demonstrated, through both theoretical analysis and extensive numerical experiments, that the automatic differentiation method outperforms the traditional finite difference differentiation in terms of efficient and accurate training of neural networks for solving PDEs. The concept of truncated entropy is introduced to quantitatively characterize the training property. Journal of Scientific Computing Volume 104: 54 (2025) DOI: https://doi.org/10.1007/s10915-025-02965-3 New Findings on CRAM and Carbon Cycling Marine dissolved organic matter (DOM) is one of Earth’s largest long-term carbon reservoirs, critical to the global carbon cycle. A research team led by Prof. HE Ding from the Department of Ocean Science at HKUST has made significant breakthroughs in understanding DOM. They identified biorefractory carboxyl-rich alicyclic molecules (CRAM) and challenged previous assumptions about their stability. Utilizing advanced ultrahigh resolution mass spectrometry, the team demonstrated that CRAM bioavailability is influenced by molecular polarity, with lower-polarity CRAM being preferentially degraded. Their findings revealed that microbial transformations produce higher-polarity CRAM, which may persist in marine environments, providing crucial insights into the complexity of marine organic carbon cycling. Science Advances Vol 11, Issue 28 DOI: https://www.science.org/doi/10.1126/sciadv.adw1148 Tensor methods for higher-order complex network statistical analysis Prof. XIA Dong from the Department of Mathematics at HKUST applies tensor-based methods to analyze large and higher-order complex networks, including multi-layer networks, hypergraph networks and temporal networks. Compared with other approaches, tensor-based methods offer better interpretability, enjoy faster computation, and preserve higher-order interactions among vertices. However, performing statistical analysis with tensor-based methods for high-order networks is challenging, particularly when the networks are extremely sparse. In his research, Prof. Xia developed sharp concentration inequalities for random sparse high-order tensors, designed analytically tractable tensor decomposition methods for high-order networks, and demonstrated the effectiveness of these methods in real-world data applications. Journal of Royal Statistical Society Series B Statistical Methodology, 2025, 00, 1–23 DOI: https://doi.org/10.1093/jrsssb/qkaf041 Email ssci@ust.hk for feedback and questions. UNSUBSCRIBE Copyright © School of Science, HKUST All Rights Reserved.