Search

Johanns and colleagues report the results (including safety, efficacy and immunogenicity) of a phase 1 clinical trial of a DNA-based personalized therapeutic cancer vaccine administered following surgical resection and radiation in patients with MGMT unmethylated glioblastoma.

In this study, the authors develop NTAC, Neuronal Type Assignment from Connectivity, using synaptic connectivity alone to identify cell types with high accuracy within minutes on a standard CPU.

The genetically encoded voltage indicators JEDI3sub and JEDI3hyp enhance monitoring of subthreshold neuronal activity in the awake mouse brain using a variety of two-photon microscopy techniques.

Upfront and lifetime costs often prevent EV adoption. Vaishnav and colleagues find that using EV batteries to shift the time of electricity purchases for other household uses can cut both owners’ electricity costs and greenhouse gas emissions.

CO and CO₂ electrolysis have progressed rapidly from laboratory test cells to pilot-scale stacks. This Comment argues that the field must now shift focus to confront the unique engineering and deployment challenges of stack-scale systems, which represent a decisive frontier for achieving commercialization and climate impact.

National climate risk frameworks catalogue threats but rarely define which risks are tolerable, to whom and how to decide. We propose embedding evaluative governance — the structures, processes and relationships that guide law, policy, deliberation, funding and evaluation — across the policy cycle to translate into accountable, equitable and adaptive climate action.

The authors use a physiological approach, finding recent extreme heat events had deadly heat stress conditions, particularly for fully exposed older people. Conditions were well below wet-bulb temperatures of 35˚C, a common heat stress indicator.

Here, the authors develop Q4ddPCR, a high-throughput assay to quantify genetically intact HIV reservoirs by targeting four regions, and demonstrate that it reduces assay dropout to 5%, tracks reservoir decay, and closely correlates with viral outgrowth.

Cardiovascular risk is driven by genes, proteins, and metabolites, yet their combined predictive value is unclear. Here, the authors develop CardiOmicScore to integrate genomics, proteomics and metabolomics and predict six cardiovascular diseases up to 15 years prior to disease onset.

Spatial targeting of HIV services may improve their effectiveness. Here, the authors use data on locations frequented by people who inject drugs in New Delhi, India to model different service prioritisation strategies.

SIDISH integrates single-cell and bulk RNA sequencing data using deep learning to identify high-risk cell populations and prognostic biomarkers, enabling in silico perturbations that could guide precision therapeutics and advance personalized medicine.

In this Review, Vasse and Velicer explore the phylogenetic and functional diversity of predators of microorganisms, conceptualizing the forms of microbial predation along gradients, including gradients of evolutionary adaptedness for predation and privatization of prey-derived nutrients.

The IDG-DREAM Challenge carried out crowdsourced benchmarking of predictive algorithms for kinase inhibitor activities on unpublished data. This study provides a resource to compare emerging algorithms and prioritize new kinase activities to accelerate drug discovery and repurposing efforts.

DNA origami enables precise nanoscale self-assembly. Authors here show how programmable control of interactions and binding angles yields curved surfaces of diverse geometries, broadening the design space for synthetic nanostructures.

Single-cell sequencing is vital for studying complex diseases but is costly. Here, authors introduce scSemiProfiler, a deep generative learning framework that infers single-cell profiles by combining bulk sequencing with single-cell data from selected samples, offering a cost-effective solution.

Graphene–quantum dots on CMOS sensor offers broadband imaging.

In this study, the authors combine interpolation techniques and deep learning methods, introducing Interpolating Neural Networks (INNs) to model complex systems with high accuracy and low computational cost.

Electrochemical reduction of carbon dioxide to more useful products is an industrially important process. Here, the authors report a nanoporous silver catalyst that efficiently and selectively reduces carbon dioxide due to its high surface area and intrinsically high activity.

Electric trucks and/or liquid fuels such as biodiesel and e-diesel could enable reductions of heavy-duty vehicle emissions and reduce reliance on fossil fuels. This Review discusses the feasibility, costs and decarbonization potential of these technologies and approaches.

Bioactivity benchmarks are key to evaluating and improving methods to predict chemicals’ specific biological activity. Here, the authors find existing benchmarks poorly suit this goal; their assays are often well-predicted simply by counting cells. They propose new guidelines and curated benchmarks.

This meta-analysis demonstrates that people who received COVID-19 vaccination had a lower risk of long COVID compared with those who were unvaccinated during the Omicron era, with booster doses providing additional protection beyond a primary course.

Diffusive light propagation represents a valuable additional tool for integrated photonic technologies. As an example, here the authors experimentally demonstrate optical equalisation of coherent light propagating in a femtosecond laser written circuit which simulates a dissipatively-coupled quantum chain.

It is desirable for scientists to be able to predict the structures, spectra and phase diagrams of molecular crystals using ab initio computation. Li et al. demonstrate such an approach, which is able to determine the phase behaviour of solid carbon dioxide at a range of pressures and temperatures.

Federated learning (FL) algorithms have emerged as a promising solution to train models for healthcare imaging across institutions while preserving privacy. Here, the authors describe the Federated Tumor Segmentation (FeTS) challenge for the decentralised benchmarking of FL algorithms and evaluation of Healthcare AI algorithm generalizability in real-world cancer imaging datasets.

This study presents an open-source Python library, GPSat, which uses Gaussian process models to interpolate satellite altimetry data. With the efficient scaling of GPSat, the authors can reconstruct complete images of high-resolution sea ice fields.

Broad-spectrum vaccines have been proposed as a tool for rapid response to emerging infectious disease threats and are in pre-clinical development. Here, the authors use mathematical modelling to assess the potential impacts of broadly protective sarbecovirus vaccines for a hypothetical “SARS-X” outbreak.

Bacteria use diverse defence systems against phages, including a 164-residue prophage-encoded protein, Rip1, which senses conserved phage assembly rings to form membrane pores that block virion maturation and trigger premature host cell death.

Surface plasmon and phonon polaritons are useful for sub-diffraction limit waveguiding of light, but phonon polaritons are advantageous in the mid-infrared. Xu et al.show that one-dimensional boron nitride nanotubes can support propagating phonon polaritons, making them a suitable platform for further study.

MK-7762 is a new oxazolidinone antitubercular agent that is structurally similar to linezolid, and demonstrated in vivo efficacy, lesion penetration and limited toxicity compared to linezolid, indicating it could be used in broad tuberculosis regimens.

An autonomous method discovers reinforcement learning rules from the cumulative experiences of a population of agents across a large number of complex environments, and the discovered rule achieves state-of-the-art performance on challenging benchmarks.

It is unclear whether trait trade-offs and optimality principles observed at the individual level scale up to the ecosystem level. Here, the authors show that plant trait coordination principles also predict patterns between community-level traits and ecosystem-scale processes.

Catherine Crichton and colleagues describe a flexible printed adhesive hydrogel bioelectrode for long-term monitoring of plant electrophysiology. By maintaining stable electrical contact for weeks, these biocompatible electrodes can transform plants into reliable living sensors.

The high mobility of graphene has the potential to enable design of devices working in a collisionless regime of carrier transport. Here the authors fabricate a ballistic nano-rectifier based on encapsulated graphene, showing intrinsic performances comparable to those of superconducting bolometers.

The attachment of a simple, unpowered, mechanical exoskeleton to the foot and ankle results in a net saving of 7% of the metabolic energy expended in human walking.

Existing studies on the economic feasibility of energy storage are system-specific without considering the decarbonisation of electricity production or impacts of GHG taxes. Here the authors applied an optimization model to investigate the economic viability of nice selected energy storage technologies in California and found that renewable curtailment and GHG reductions highly depend on capital costs of energy storage.

Power grids often fail during extreme weather events such as hurricanes, leaving millions of customers without electricity. A large-scale analysis of the operation of power grids in an extended geographical area now reveals that such events exacerbate vulnerabilities that are obscured during normal operation.

Nanoplastics (NPs) threaten water safety as conventional coagulants fail to remove them. Here, authors develop an in-situ Fe(III) method that forms encapsulating nanosheets, enabling efficient NP removal with stronger anti-interference in natural waters.

How is specific ribosomal RNA expression precisely controlled in the malaria parasite? Dr. Couble et al. show that a switch between distinct energy-generating pathways during parasite development directly impacts the accessibility to rDNA chromatin.

Variants in the PSMC5 gene impair proteasome function and cellular homeostasis, altering brain development in children. This study reveals underlying molecular mechanisms contributing to this neurodevelopmental phenotype, and suggests therapeutic leads for neurodevelopmental proteasomopathies.

mRNA vaccines targeting SARS-CoV-2 also sensitize tumours to immune checkpoint inhibitors.

Aerial surveys over the Permian Basin found 500+ major methane leaks, many recurring. A few sites leaked continuously and offer quick mitigation wins. These super-emitters may produce ~50% of regional emissions, underscoring the need for frequent monitoring.

By optimizing the timing of electricity purchases for electric vehicle (EV) charging at home, as well as shifting electricity purchases for other household loads, US EV owners could reduce their lifetime charging costs by 40–90%, and lifecycle greenhouse gas emissions from household electricity use by 70–250%. Integrating EVs with homes can increase the greenhouse gas reductions they deliver, while reducing the cost of EV ownership.