Browse Articles
Discover research articles across all indexed journals
Vertical nanodiamond dominated sheets possessing both high capacitance and high n-type Hall mobility
Understanding pre-training data effects in retinal foundation models using two large fundus cohorts
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.
Doublet microtubule-associated tektins and enzymes differentially regulate sperm flagellar integrity and motility
Abstract Doublet microtubule (DMT)-associated proteins assemble and drive sperm flagella, which are essential for successful fertilization. However, the exact roles of different DMT-associated proteins in sperm function and the underlying molecular mechanisms remain elusive. Here, we generate four gene-knockout mice based on high-resolution structures targeting distinct DMT components: two intermediate filament-like tektins (TEKT1, TEKT5) and two enzymes (TSSK6, DUSP21). The depletion of TEKT1, shared by sperm flagella and motile cilia, causes male infertility characterized by impaired sperm motility and loss of the tektin bundle, whereas sperm-specific Tekt5 knockout (KO) mice remain fertile with largely normal flagellar function, indicating functional divergence within the tektin family. Tssk6 KO spermatozoa exhibit severely disturbed morphology and motility, resulting in homozygote infertility and heterozygote subfertility. Phosphoproteomics reveals dysregulated phosphorylation of axonemal proteins, highlighting the critical role of kinase-mediated signaling in regulating sperm motility. Conversely, Dusp21 KO mice display no fertility or sperm motility defects, suggesting compensatory phosphatase activity. Phenotypic comparisons between Tekt1 and Tssk6 KO mice suggest their involvement in distinct subtypes of asthenozoospermia. Overall, this study elucidates how filamentous and enzymatic DMT proteins govern sperm function through divergent mechanisms, which have implications for molecular diagnosis of male infertility.
Role of CTGF-LRP1 in impaired healing of cesarean section incisions
Data storage and retrieval with unnatural proteins expressed via E. coli
Attention-related modulation in the superior colliculus encodes perceptual sensitivity, but not perceptual choice
Rapid synthesis of micron-thick flexible graphite films via non-equilibrium carbon flux engineering
Abstract The scalable synthesis of high-quality graphite materials remains a formidable challenge due to the inherent trade-off between crystalline perfection and manufacturing efficiency. Existing forms of graphite, such as highly oriented pyrolytic graphite (HOPG) and Kish graphite, suffer from sluggish pyrolytic processes, limited carbon diffusion rates and energy-intensive protocols, often requiring several days for production. Here, we report a pulsed Joule heating-induced carburization (PJHIC) strategy that exploits transient non-equilibrium states to enable rapid carbon diffusion and segregation in metal substrates. By applying instantaneous thermal shocks ( > 1300 °C, > 300 °C/s heating rate) to solid carbon precursor-coated nickel and cobalt foils, we demonstrate the rapid carbon transport in bulk metals and achieve a vertical graphite growth rate of 730 nm/min, which is an order of magnitude faster than conventional methods. Cyclic temperature pulses further enable the synthesis of 1–5 μm-thick ABA-stacked graphite films with millimeter-scale grain sizes. The resulting rapid epitaxially grown graphite films exhibit a highly ordered crystalline structure and exceptional thermal conductivity (1314 W m –1 K –1 ), comparable to high-quality HOPG and Kish graphite. This work establishes a non-equilibrium synthesis paradigm for high-quality layered materials, bridging atomic-scale precision with industrial-scale manufacturing.
Probabilistic day-ahead forecasting of system-level renewable energy and electricity demand
Abstract Increasing shares of wind and solar generation, together with rising electricity demand, introduce growing uncertainty into power system operations. Accurate day-ahead forecasts of electricity demand and renewable generation are essential for system operators to coordinate electricity markets and maintain reliability at low cost. Here, we show that forecasting based on joint probability distributions of demand and renewable supply can substantially improve system-level forecasting performance using publicly available weather data. We develop multiple day-ahead forecasting models that combine machine learning methods to identify relevant weather variables with probabilistic approaches to quantify forecast uncertainty, and we evaluate these models using proper scoring rules. Applied to the three zones of the California Independent System Operator, the best-performing model improves forecast skill by 25% relative to current benchmarks. We further show that forecasts based on joint probability distributions enable a more effective allocation of operating reserves than conventional deterministic approaches, highlighting the potential of probabilistic machine learning to enhance market efficiency and grid stability in increasingly decarbonized power systems.
Hybrid macrophage-mitochondria extracellular vesicles for mitochondrial ROS regulation in diabetic wounds
Chromatin-intrinsic mechanisms determine orientation-specific class switch recombination
Topology-controlled dynamic conjugated oligomers from tetra-arylsubstituted alkene building blocks
Adaptive optical waveguide system for large-area and overheating-preventing phototherapy in deep tissue
Abstract Phototherapy, valued for its non-toxicity, selectivity, and minimal trauma, is predominantly applied to treat superficial diseases due to the limited penetration of light through tissues. While optical-fiber-assisted interventional phototherapy addresses this limitation, it lacks an immediate mechanism to mitigate overheating of surrounding healthy tissues. To improve the biosafety of interventional phototherapy, we develop an adaptive optical waveguide system (AOWS) based on a negative feedback modulation mechanism. The AOWS employs a thermally responsive liquid as the waveguide gating material, characterized by a precisely tunable low critical solution temperature (LCST). When the temperature surpasses the LCST, one-dimensional light propagation is scattered, providing effective thermal regulation. Furthermore, the design supports adjustable optical fiber outlets, with convex or concave configurations and varied curvatures, enabling precise control of the divergence angle. Superior to conventional optical fibers with smaller divergence angles, the AOWS facilitates closer placement to the lesion site, delivering a larger illumination area while significantly reducing the light pathway through normal tissue. More importantly, it provides thermal protection almost like an “on-off” switch without relying on irradiation power, ensuring enhanced safety and efficacy.
Research on the socio-spatial resilience evaluation and evolution of the central area of Chengdu in transitional China
3D-printed core–shell scaffolds with a biphasic calcium phosphate core and GelMA hydrogel shell for bone tissue engineering
A comparative anti-proliferative and immunomodulatory analysis in wild and lab-acclimatized seaweed extracts unravel the functional biopotentials of Acrosiphonia orientalis
Bone marrow stromal cells enhance chondrocyte function and autophagy via mTOR signaling
Immediate effects of real time feedback and kinesiotaping on kinematics and muscle activity in athletes with dynamic knee valgus
Microstructural refinement and mechanical property enhancement of AZ91 magnesium alloy via room-temperature multi-directional forging
XMD8-92 and JWG-045 exhibit anti-ferroptotic activities, independently of inhibiting ERK5
Abstract Extracellular-regulated protein kinase 5 (ERK5) is an emerging therapeutic target in cancer, and small-molecule ERK5 inhibitors have been widely employed to define its role in tumour biology. Here, we show that the commonly used ERK5 inhibitors XMD8-92 and JWG-045 suppress RSL3-induced ferroptosis in breast cancer cells, in contrast to the next-generation ERK5 inhibitors JWG-071 and BAY-885, and the MEK5 inhibitor BIX02189. Using CRISPR-mediated gene editing, we generated ERK5-deficient breast cancer cells and found that XMD8-92 and JWG-045 retained their anti-ferroptotic activity against RSL3 in the absence of ERK5 expression, indicating clear off-target effects. Pathway-level analysis of bulk RNA-sequencing data using FerrDb-curated gene sets revealed no global alteration in ferroptotic activity in BT474 cells following XMD8-92 treatment. Interestingly, XMD8-92 did not inhibit RSL3-induced lipid peroxidation and preserved cell viability even after RSL3-induced ferroptosis initiation. Based on these observations, we propose that XMD8-92 confers transient resistance to ferroptotic cell death by maintaining plasma membrane integrity, potentially through enhanced membrane repair mechanisms. Collectively, these findings reveal a previously unrecognised off-target anti-ferroptotic activity of XMD8-92 and JWG-045, further highlighting the limitation of these compounds for ERK5-specific mechanistic studies.