Search

AlphaFold’s success in protein structure predictions has led to similar attempts to predict interactomes. Here, the authors demonstrate that AI-based screens are very limited in discovering truly novel interactions compared to experimental screens, exposing open challenges in interaction prediction.

Xu and Wang et al. have developed a mouse-adapted Yezo virus model that recapitulates human disease and demonstrates that the broad-spectrum antiviral ribavirin, but not remdesivir, protects against lethal infection.

The authors present a multi-core neuromorphic architecture that enables energy‑efficient training of deep spiking neural networks via backpropagation. The design features a multi-engine core, multi-level parallelism, sparsity optimization, and FPGA validation.

Zinc-based flow batteries offer safe, low-cost energy storage but suffer dendrite-related failure. Here, authors employ quantum sensing to link local temperature gradients to dendrite growth and develop a liquid metal electrode that mitigates hotspots, enabling long-term stable cycling.

Robustness checks and reproduction of analyses with existing and updated data based on 110 articles in economics and political science journals with data and code-sharing requirements found high levels of robustness and reproducibility and determined that robustness was not dependent on author characteristics or data availability.

Host cell protein degradation machinery including the lysosome is involved in the processing and degradation of protein antigens. Here the authors develop lysosomal targeting vaccine approaches and test these in murine models of influenza infection.

Micro-batteries power miniaturized electronics, but most are limited by single-cell reactions. Here, the authors report an in situ conductivity-enhanced dual-reaction strategy that enables sequential reactions in Zn | |Bi2O3@Ag2O micro-batteries, delivering higher capacity, energy density, and power density.

Wu et al. identify circulating proteins associated with cardiovascular and neurological imaging traits and diseases, and validate GDF15 as a shared determinant of heart and brain pathologies in mice.

A new Matteson-type homologation method inserts carbon units into boronic esters using electrochemistry.

Tuning the interfacial structure of silver nanoclusters via steric hindrance modulates their photoluminescence. Locking the ligand shell endows the clusters with room-temperature photoluminescence, thereby enabling atomic-level design of optical devices.

Electrochemical nitrate reduction can simultaneously remove pollutants and produce ammonia but is often limited by hydrogen side reactions. This study shows that nanoconfined CuCo catalysts enable a proton-coupled electron transfer pathway that delivers highly selective, efficient and stable ammonia synthesis.

Annunziato, Quan and Donckele et al. identify G3BP2 (Ras–GAP SH3 domain-binding protein 2) as a molecular glue-induced neosubstrate of the CRL4^CRBN E3 ubiquitin ligase. The CRBN–glue neosurface uses a molecular surface mimicry mechanism to recruit and degrade G3BP2 in a compound-dependent manner.

A platform combining site-specific probe orientation with scalable In₂O₃ field-effect transistors enables attomolar, serum-compatible detection of endometrial cancer biomarkers, offering a portable and highly sensitive foundation for accessible point-of-care cancer diagnostics.

NatD is an acetyltransferase responsible for N-α-terminal acetylation of the histone H4 and H2A and has been linked to cell growth. Here the authors show that NatD-mediated acetylation of histone H4 serine 1 competes with the phosphorylation by CK2α at the same residue thus leading to the upregulation of Slug and tumor progression.

An in-sensor wireless computing architecture integrates imaging, compression and transmission into a single step, dramatically reducing latency for real-time remote-sensing applications.

He et al. identify a central regulator of condensate size and dynamics in the Chlamydomonas CO₂-fixing condensate, the pyrenoid. KEY1 phosphorylates EPYC1, a linker protein for Rubisco, to maintain pyrenoid size and number through the cell cycle.

Spent LiCoO₂ batteries pose challenges for sustainable recycling. Here, the authors develop a flash Joule heating method to upcycle spent LiCoO₂ into a sulfur-stabilized Li₆CoO₄ lithium donor, enabling efficient lithium replenishment and improving full-cell durability.

A single-step strategy is reported for the preparation of delaminated MXenes with positively charged surfaces from MAX phases. By combining Lewis acid etching with gas-mediated delamination in deep eutectic solvents, this approach offers a green, scalable and efficient route for MXene synthesis.

A meta-analysis of plant nuclear genomic data provides fundamental insights into the genomic nature of differences between plant species and informs future strategies for DNA-based identification of plants using nuclear genomes.

A three-band time-multiplexed photonic quantum walk simulates a Floquet non-Abelian topological insulator. The observation of the anomalous non-Abelian phase confirms the bulk-boundary correspondence using the topological configuration of phase-band singularities.

Phospholipid scrambling at the endoplasmic reticulum is essential for establishing a dynamic equilibrium to maintain cellular homeostasis.

Estimates from the Global Burden of Disease study reveal declines in chronic respiratory disease mortality between 1990 and 2023—especially during the COVID-19 pandemic—but the burdens on people affected remain substantial.

In this study, authors reveal an evolutionary strategy in bacteria, where multidrug resistance are stably maintained through multicopy heterogeneous genes on a single plasmid, conferring an adaptive advantage under fluctuating clinical drug pressures.

Despite various advantages of aluminium batteries, their application is hindered by the use of ionic liquid electrolytes. Here this work shows an organic formulation that is non-corrosive and delivers better performance, opening vast chemical space for electrolyte design.

This study reports a post-assembly, reversible crosslinking strategy that enhances lipid nanoparticle (LNP)-mediated mRNA delivery while preserving efficient intracellular release. The resulting crosslinked LNPs enable improved endosomal escape, sustained in vivo expression and robust immune and antitumor responses across multiple clinically relevant LNP platforms.

Alphaherpesvirus infections pose a health challenge due to the lack of effective broad-spectrum antiviral strategies. Here, the authors identify a broadly neutralizing antibody, 16F9, which targets glycoprotein B, offering cross-protection against multiple viruses and demonstrating robust efficacy in mouse models, paving the way for antiviral interventions.

Converting carbon dioxide into valuable multi-carbon products remains challenging under acidic conditions. Here, the authors develop a covalent organic framework–modified copper nitrate catalyst that enables stable ethylene production at high current densities.

A divergent electrochemical nitrogen atom insertion into saturated carbocycles is reported, enabling selective access to either functionalized quinolines or N-alkylated saturated N heterocycles. This approach is applicable to late-stage functionalization and the synthesis of bioactive molecules.

Quantitative analysis of polymer LED degradation under current stress provides insight on the role of hole traps and their formation. Blending of the emitting material with large-bandgap semiconductors leads to trap dilution and improved stability.

The authors present a 3.5-million-year-long pollen record from the Zoige Basin of the eastern Tibetan Plateau, 3,442 m above sea level. The ~5,000 pollen assemblages retrieved from the core reveal many ecosystem transitions during this time and, when correlated with climatic curves, indicate what effects future warming may have on regional vegetation.

A ductile inorganic-rich solid–electrolyte interphase that retains its structural integrity and allows easy ion diffusion enables a long cycle life under practical conditions for lithium metal batteries.

Maximizing metal-substrate interactions through self-reconstruction is a key strategy for efficient oxygen evolution catalysts. Here, the authors report high-density iridium atoms embedded in ultrathin oxyhydroxide nanosheets, showing high performance in the oxygen evolution reaction.

Affinity-proteomics platforms often yield poorly correlated measurements. Here, the authors show that protein-altering variants drive a portion of inter-platform inconsistency and that accounting for genetic variants can improve concordance of protein measures and phenotypic associations across ancestries.

Directed evolution remains largely untapped for probiotic optimization, despite its promise for human health maintenance and disease therapy. Here the authors evolve B. lactis in vivo in germ-free mice for enhanced bile acid-metabolizing capacity via a high-fat, high-cholesterol diet.

An approach that integrates electrochemical CO₂ reduction and biological denitrification provides a new solution for nitrogen removal, where the formate obtained from CO₂ is used directly as the carbon source in the denitrification process in wastewater treatment.

Yang et al. compare neoadjuvant triple therapy versus direct hepatectomy in patients with resectable intermediate or advanced hepatocellular carcinoma. The findings indicate that the neoadjuvant regimen improves overall and event-free survival but is associated with a higher risk of serious adverse events and postoperative complications.

Spatial transcriptomics (ST) technologies link tissue morphology with gene expression, but remain expensive to use; furthermore, models that predict ST data from histopathology images possess considerable limitations. Here, the authors develop STimage, a deep learning probabilistic framework for ST prediction from histopathology images while prioritising robustness and interpretability.

Hypergraph-based modeling of metabolic networks integrates chemical structure and topology to accurately predict missing reactions, enabling efficient gap-filling and improved phenotype prediction in genome-scale metabolic models.

Prior hypoglycemia alters the paracrine interaction between islet α and δ cells, leading to impaired counter-regulatory glucagon secretion through somatostatin hypersecretion, increasing the risk of recurrent hypoglycemia.

Here, authors use a membrane electrode assembly is conducted to produce adipic acid with high faradaic efficiency and productivity at an industrially relevant current density, while achieving >50 hours stability.

Signalosomes amplify cellular responses, but how they avoid uncontrolled activation is unclear. The authors use quantitative super-resolution imaging and mathematical modeling to reveal optimal RIP3:RIP1 ratios that ensure precise cell death signaling.

Emu3 enables large-scale text, image and video learning based solely on next-token prediction, matching the generation and perception performance of task-specific methods, with implications for the development of scalable and unified multimodal intelligence systems.

Current battery recycling processes face sustainability challenges. Using gas evolution in water electrolysis, this work realizes fast separation of active electrode materials from current collectors before their dry refabrication for electrodes without compromising performance.

Reduced-dimensional halide perovskites are promising for light-emitting diodes but suffer from photo-degradation. Here Quan et al. identify the edge of the perovskite nanoplatelets as the degradation channels and use phosphine oxides to passivate the edges and boost device performance and lifetime.