Search

When 100 social and behavioural science claims were examined, 34% of reanalyses closely matched the original results, with 74% reaching the same conclusion, revealing limited robustness of single-path analyses and the need to address analytical uncertainty.

CREsted is an efficient and user-friendly toolbox for analysis, modeling and design of cell-type-specific enhancers across diverse species.

ATF6α activation in human and preclinical models of hepatocellular carcinoma is significantly associated with an aggressive tumour phenotype characterized by reduced survival, glycolytic reprogramming and local immunosuppression.

Generating high-quality training data for machine learning is costly. Here, authors include sequence-to-function modeling in benchmarking of custom and commercial droplet-based scATAC platforms, and release a new Drosophila embryo atlas along with a new mouse cortex atlas, assessed for model interpretability.

Humanity’s Last Exam, a multi-modal benchmark at the frontier of human knowledge, is designed to be an expert-level closed-ended academic benchmark with broad subject coverage.

A platform using matched patient-derived lung tumouroids and healthy lung organoids enables accurate examination of patient responses to CAR T therapy and offers a faithful framework for improved CAR T design.

A cavity-array microscope is realized using intra-cavity lenses to create a two-dimensional array of over 40 modes, each coupled to a single atom in free-space.

Neuromorphic photonic systems can incur significant energy for moving and converting data between digital and analog domains. This work shows that integrating analog memory into these processors can save 26 × power over conventional digital-to-analog architectures while keeping  > 90% inference accuracy.

Single-cell RNA sequencing efforts have made large amounts of data available for transcriptomics research. Simon and colleagues develop a neural network embedding approach that avoids batch effects, such that it can rapidly and efficiently integrate large datasets from different studies.

Aligning foundation models with human judgments enables them to more accurately approximate human behaviour and uncertainty across various levels of visual abstraction, while additionally improving their generalization performance.

The rapid expansion of agricultural irrigation raises concerns about exacerbating water scarcity, but land–atmosphere interactions are often overlooked. This study isolates irrigation impacts from other drivers using a multi-model framework to reveal that historical irrigation expansion substantially reduces net atmospheric water influx, intensifying drying trends and accelerating terrestrial water storage depletion, urging immediate mitigation strategies.

A doped quantum antiferromagnet is obtained by using a Rydberg tweezer array comprising three levels encoding spins and holes to implement a tunable model that allows the study of previously inaccessible parameter regimes.

Natural products have historically made a major contribution to pharmacotherapy, but also present challenges for drug discovery, such as technical barriers to screening, isolation, characterization and optimization. This Review discusses recent technological developments — including improved analytical tools, genome mining and engineering strategies, and microbial culturing advances — that are enabling a revitalization of natural product-based drug discovery.

A scalable solution involving direct wafer-bonding of high-quality, epitaxially grown gallium phosphide to low-index substrates is introduced. The promise of this platform for integrated nonlinear photonics is demonstrated with low-threshold frequency comb generation, frequency-doubled combs and Raman lasing.

By unifying data from engineered β-barrel nanopores and supported by modelling, it is demonstrated that the lumen charge in a β-barrel nanopore governs rectification and voltage-driven gating, with applications in computing using nanofluidic synapses.

Spinal cord injury leads to lasting motor and sensory loss, with limited treatment options. Here, the authors show that subdural electric field stimulation delivered across the injury via an ultra-thin implant improves functional recovery in rats.

Polygenic scores for body mass index and obesity constructed using data from 5 million people with different ancestries were associated with distinct weight gain trajectories from early childhood to adulthood.

Risk stratification in non-small cell lung cancer (NSCLC) remains challenging. By combining multiplex immunofluorescence, H&E histology, and AI, the study identifies spatial “cell-niche” patterns that enhance survival prediction beyond UICC8 staging. These patterns reclassify many stage I patients as high risk, revealing potentially undertreated cases and establishing spatial tumor microenvironment features as clinically actionable biomarkers.

Second-generation oxidation boosts the production of anthropogenic organic aerosol, according to an analysis of urban haze in an experimental chamber.

A large genome-wide association study of more than 5 million individuals reveals that 12,111 single-nucleotide polymorphisms account for nearly all the heritability of height attributable to common genetic variants.

A genome-wide association meta-analysis study of blood lipid levels in roughly 1.6 million individuals demonstrates the gain of power attained when diverse ancestries are included to improve fine-mapping and polygenic score generation, with gains in locus discovery related to sample size.

A phase 1 clinical trial provides evidence that a vaccine against mutant IDH1 is safe and produces a T helper immune response in patients with glioma.

This work highlights the technical issues in previous approaches and introduces a preprocessing approach along with a software package, MethSCAn, for single-cell bisulfite sequencing data analysis.

Substrate-specific proteases have an enormous potential in the life sciences, but tailoring their specificity remains challenging. Here, the authors describe a data-driven approach combining DNA recording and epistasis-aware deep learning to augment protease specificity engineering at large scale.

The authors provide cryo-EM reconstructions of human transfer RNA splicing endonuclease during substrate recognition, rationalizing disease-relevant mutations.

Feuerer and colleagues use multiomic analyses to identify DNA methylation, chromatin accessibility and gene expression programs that govern the tissue adaptation of human skin T_reg cells.

EU wine geographical indications are bound to a strict regulatory system, limiting their potential to adapt to climate change. In this study, analysis of social, environmental and economic characteristics of over 1000 EU wine regions suggests that innovation and flexibility are key to increase the resilience of European viticulture.

53BP1 is marker of double-strand breaks and accumulates in nuclear foci. These foci are shown to accumulate in G1 at lesions generated by replication stress and may shield lesions from erosion.

The spatial architecture of multiple myeloma remains to be explored. Here, the authors perform bulk and single cell sequencing for samples from newly diagnosed patients and reveal gene signatures associated with focal lesions and spatial heterogeneity in the tumour microenvironment.

The need for redox balancing limits the array of fermentable substrate-product combinations in anaerobic microbe-based bioproduction. Here, the authors design and engineer an E. coli strain with new-to-nature aerobic fermentative metabolism that allows tightly controlled re-balanced fermentations.

An antibody–drug conjugate that targets the pan-haematopoietic marker CD45 combined with transplanted stem cells engineered to be shielded from it can eradicate leukaemic cells while preserving haematopoiesis.

Climate models, impact models and demographic data are used to estimate the number of people projected to experience unprecedented lifetime exposure to extreme climate events across multiple dimensions, including birth year, warming scenario and vulnerability.

Hofer et al. show that fasting promotes the synthesis of spermidine, which stimulates eIF5A hypusination to induce autophagy and increase lifespan in various species in a conserved manner.

Single-cell analysis of the transcriptome, chromatin accessibility and methylome of adult neural stem cells and astrocytes demonstrates that stemness is driven by methylation profiles distinct from those of astrocytes.

I-Motifs (iM) are non-canonical DNA structures potentially forming in the accessible, single stranded, cytosine-rich genomic region, but the specific contributions of several factors involved in their formation are unknown. Using in-cell NMR, the authors examined DNA i-motif formation in human cells at body temperature, suggesting i-M occur in a small portion (<1%) of genomic sites predisposed to its formation.

Loss of let-7, a key microRNA, causes lung alveolar stem cells to grow abnormally and dieprematurely, leading to scarring and lung damage. Here the study links let-7 loss to cancer-like signals and epigenomic gene changes, revealing new treatment paths for lung fibrosis.

HERC2 regulates the retention of repair proteins 53BP1, RAP80 and BRCA1 on damaged chromosomes in response to ionizing radiation by forming a complex with the ubiquitin ligases RNF8 and RNF168. Hecr2 deficiency leads to radiosensitivity.

Single-cell RNA sequencing is a powerful method to study gene expression, but noise in the data can obstruct analysis. Here the authors develop a denoising method based on a deep count autoencoder network that scales linearly with the number of cells, and therefore is compatible with large data sets.

Phosphatidic acid constitutes less than 1% of total lipids, but does not play a minor role. It serves as a lipid precursor, affects membrane curvature and regulates signalling transduction. Now, by employing a new powerful PAleon biosensor, the researchers visualized PA dynamics in different pH conditions corresponding to salt stress responses in Arabidopsis.

Zeiser and colleagues show that CAR T cell therapy results in upregulation of the TGFβ-activated kinase-1 (TAK1)–NF-κB–p38 MAPK pathway in microglia, causing neurocognitive defects, and find that TAK1 inhibition can reduce immune effector cell-associated neurotoxicity syndrome.

Assessing person-level human brain maps across 18 fMRI studies, the authors identify separable representations of pain, cognitive control, and negative emotion in the medial frontal cortex that generalize across different studies and tasks.

The oxidation of CO to CO₂ on rutile TiO₂ is a model reaction in heterogeneous catalysis, but the effect of substrate-specific oxygen activation timescales remains underexplored. Here, the authors monitor the reaction through time-resolved photoelectron spectroscopy at a free-electron laser and propose an O₂-TiO₂ charge transfer complex responsible for initiating the oxygen activation pathway for CO oxidation on rutile, showing that despite anatase being the more active photocatalyst, oxidation dynamics on rutile are faster.

A direct imaging study demonstrates the existence of a giant planet in a wide orbit around the high-mass b Centauri binary system, and uses measurements of the orbital properties to discuss its formation mechanism.

Rinnerthaler et al. perform a randomized phase 2 trial of neoadjuvant atezolizumab in combination with dual HER2 blockade plus epirubicin in patients with early HER2-positive breast cancer and report positively on efficacy and safety.

Development of the secondary palate is a complex process. Here, the authors profile mouse palatogenesis through single-cell multiome sequencing, revealing dynamic gene regulation across embryonic days (E) 12.5, E13.5, E14.0, and E14.5.

The hippocampus in mammalian brain varies in size across individuals. Here, Hibar and colleagues perform a genome-wide association meta-analysis to find six genetic loci with significant association to hippocampus volume.

The human endoderm-derived organoid cell atlas (HEOCA) presents an integrative analysis of single-cell transcriptomes across different conditions, sources and protocols. It compares cell types and states between models, and harmonizes cell annotations through mapping to primary tissues.