Search

A consensus cell type atlas for the fly brain provides both an intellectual framework and open-source toolchains for brain-scale comparative connectomics.

Mendelian randomization (MR) identifies causal relationships from observational data but has increased error rates when the genetic variants used as instruments come from a single region, a typical scenario when assessing molecular traits like protein or metabolite levels as risk factors. Here the authors introduce a single-region pleiotropy-robust MR method, validating the method on three ground truth sources, showing its capability to identify disease-causing molecular traits.

LINE-1 activity was quantified in a large, pan-cancer dataset, finding locus-specific heterogeneity and new associations using a computational pipeline. A mathematical mediation model of p53 and L1 interactions was inferred. Somatic retrotransposition was seen in Li-Fraumeni Syndrome with heritable TP53 mutations.

This study assessed COVID-19 social science preprints’ replicability using structured groups. Both beginners and more-experienced participants used a elicitation protocol to make better-than-chance predictions about the reliability of research claims under high uncertainty.

Using metabolic dysfunction-associated steatohepatitis-driven hepatocellular carcinoma mouse models, an ATP citrate lyase inhibitor reduces tumour burden and enhances efficacy of current standards of care.

Both environmental factors and genetic factors influence human immunity. Albert and colleagues leverage data from the Milieu Intérieur Consortium to comprehensively describe the effects of lifestyle, environment and genetics on human innate and adaptive immunity.

BindCraft, an open-source, automated pipeline for de novo protein binder design with experimental success rates of 10–100%, leverages AlphaFold2 weights to generate binders with nanomolar affinity without the need for high-throughput screening.

Medulloblastomas (MBs) are highly heterogeneous paediatric brain tumours that remain challenging to treat. Here, the authors integrate proteomics, epigenomics, transcriptomics and post-translational modification analyses to find molecular subgroups and potential therapeutic targets in MB tumours.

Despite their differences, the rarer sarcoma CIC::DUX4 sarcoma (CDS) is typically treated with therapies developed for Ewing Sarcoma (EwS) with limited success. Here, the authors develop a co-clinical drug response profiling platform to establish patient-derived CDS and EwS tumoroids, identifying MCL1 inhibition as a promising therapeutic approach in CDS.

The effects of different choices on preprocessing pipelines for functional connectomics remain unclear. Here, the authors systematically evaluate a multitude of pipelines on resting-state fMRI, revealing a number of optimal pipelines for functional brain network analysis.

In people with severe brain injuries, stimulation restored consciousness by engaging a deep brain circuit for wakefulness—revealing a target that may also guide treatment in stroke and epilepsy.

COMET, an artificial intelligence method that improves the analysis of small medical studies using large clinical databases, has been created. COMET can help develop better artificial intelligence tools and identify key biomarkers across many diseases, potentially changing medical research.

KDM4C regulates cathepsin L-mediated histone H3 N-terminal tail clipping through grainyhead-like 2 (GRHL2) methylation in breast cancer. This link between the cellular redox state and chromatin remodeling might represent a therapeutic target in KDM4C-amplified tumors.

Targeting histone deacetylases (HDACs) alone has shown limited success in solid tumours. Here, authors report that the HDAC1/2 inhibitor romidepsin confers responsiveness to receptor tyrosine kinase inhibitors, with enhanced therapeutic effects in models of hepatocellular carcinoma, leading to tumour regression and an immune-stimulatory profile.

The social exposome—lifelong social and economic adversity—can shape brain health and dementia risk. Here, the authors show that an adverse social exposome is linked to poorer clinical, cognitive, and brain changes in Latin American older adults.

The functions of small nucleolar RNAs (snoRNAs) in cancer remain to be studied. Here the authors show that the snoRNA Snord67 promotes lymph node metastasis of breast cancer cells through guiding modifications of the spliceosome and regulating the splicing landscape.

Neutrophils actively induce tumour necrosis, driving vascular occlusion, pleomorphic necrosis and metastasis.

Statistically unbiased prediction utilizing spatiotemporal information in imaging data (SUPPORT) is a self-supervised deep learning approach to accurately denoise voltage and calcium imaging data while preserving true dynamic signals.

Persisting symptoms after concussion (PSaC) present a complex neuropsychiatric challenge with limited treatment options due to inconsistent neuroimaging findings. Here the authors employ a multi-analytic approach to identify the salience network as a core dysfunction hub in PSaC, proposing specific cortical regions as potential targets for personalized neuromodulation therapies.

This multiomic study, including single-nucleus DNA methylation and chromatin conformation matched with single-nuclei RNA sequencing, provides insights into the epigenomic landscape of human subcutaneous adipose tissue.

Here the authors dissect the developmental and functional relationship between tumor-responsive cytotoxic T cells in the tumor versus the tumor-draining lymph nodes (tdLNs), finding that stem-like T_PEX cells dependent on MYB in the tdLNs are required for CD8⁺ T cell tumor infiltration and ICB responses.

Use of artificial intelligence to mine proteomes of archaea led to the discovery of archaeasins, antimicrobials that kill drug-resistant bacteria in laboratory and animal models, offering a promising source of future antibiotics.

This RCT finds that providing information and support to target cognitive and behavioural barriers eliminates early childcare application gaps for low-income and immigrant families in France. While application rates increased, the impact on access rates for low-SES and immigrant households was limited.

Animals alternate between active periods and periods of rest or sleep. This study in fruit flies points to brain metabolism as a cause for this and shows that a network of glial cells interacting with neurons links brain function with the need for rest and sleep.

This study explores the genomic and transcriptomic landscapes of triple-negative breast cancer in African American women. The authors show that the mutational profile is broadly similar to that observed in European and East Asian ancestry women while highlighting some interesting differences.

Malonyl-CoA (M-CoA) is essential for polyketide biosynthesis, but its limited availability constrains production. Here the authors engineer and evolve an orthogonal M-CoA pathway in Escherichia coli to improve M-CoA metabolism, increasing M-CoA levels and polyketide yields.

An artificial intelligence model defines a data-driven set of total parenteral nutrition compositions to assist clinicians in personalized treatment of neonates in intensive care and is able to adapt recommendations to patient status, with validation from large external cohorts and a blinded reader study.

Decoding the content of extracellular vesicles (EVs) is key to understanding cell communication. Nikonorova et al. used targeted proximity labeling to uncover polycystin EV cargo and cargo-loading mechanisms in C. elegans.

The number of individuals in a given space influences animal interactions and network dynamics. Here the authors identify general rules underlying density dependence in animal networks and reveal some fundamental differences between spatial and social dynamics.

Wastewater treatment plants are important reservoirs of antibiotic resistance genes (ARGs). Here, the authors analyze ARGs in a global collection of samples from wastewater treatment plants across six continents, providing insights into biotic and abiotic mechanisms that appear to control ARG diversity and distribution.

Experiments in mice show that a cortico-thalamic circuit generates prediction-error signals in primary visual cortex that amplify visual input that deviates from animals’ expectations.

Understanding phylogenetic relationships among species is key to studying evolutionary transitions, but the growing scale of sequence data poses challenges for current methods. This study presents PhyloTune, a fine-tuning approach using DNA language models to enable efficient phylogenetic updates.

Deep learning can be used to mine the proteomes of all available extinct organisms for the discovery of compounds with antibiotic properties.

Proteins are often modified by complex carbohydrates (N-glycans). Here, authors identified gene regulators of this process and uncovered links between plasma protein Nglycosylation, metabolic and liver diseases, and anti-inflammatory proteins.

CellSpectra identifies functionally relevant changes in cellular pathways when analyzing single samples in comparison to a reference atlas, here applied to rodent models and patient samples of kidney disease.

Perceptual abilities can be improved by training, up to certain limits. Martin et al. show that vagus nerve stimulation in mice boosts performance on an auditory task via cholinergic modulation, beyond the level achieved by training alone.

A large brain imaging study of over 2000 people worldwide shows that fear conditioning engages brain regions linked to emotion and attention, with differences in individuals with anxiety or depression and strong influences from task design.

Children from lower socioeconomic backgrounds often show lower academic achievement, commonly linked to limited resources. Here, the authors show that reduced exploration–a behavior tuned for learning in childhood–helps to explain these learning gaps.

Mathematical models of V1 seek to explain the response properties of V1 neurons, often with more complex models providing more accurate predictions. Here, the authors show that deep neural network models of mouse and monkey V1 can be dramatically simplified to a two-layer “minimodel" while retaining high accuracy.

The original Cancer Cell Line Encyclopedia (CCLE) is expanded with deeper characterization of over 1,000 cell lines, including genomic, transcriptomic, and proteomic data, and integration with drug-sensitivity and gene-dependency data.

A survey across 90 societies reveals that variation and change in everyday norms are explained by a single value dimension: the priority societies place on individualizing versus binding moral concerns.

Individual variation in fMRI-derived brain networks is reproduced in a model using only the smoothness (autocorrelation) of the fMRI time series. Smoothness has implication for aging and can be causally manipulated by psychedelic serotonergic drugs.

Cells often navigate through confined spaces. Now it is shown that cells retain a mechanical memory of previous confinement events, which makes them more efficient at migrating through narrow microenvironments.

Stratified medicine promises to tailor treatment for individual patients, however it remains a major challenge to leverage genetic risk data to aid patient stratification. Here the authors introduce an approach to stratify individuals based on the aggregated impact of their genetic risk factor profiles on tissue-specific gene expression levels, and highlight its ability to identify biologically meaningful and clinically actionable patient subgroups, supporting the notion of different patient ‘biotypes’ characterized by partially distinct disease mechanisms.

Complement proteome engagement is strongly linked to kidney outcomes in diabetes. This translational study leveraged five cohorts of over 4,500 person-years and high-throughput proteomics to enable potential biomarker-guided drug development.

Soft objects challenge visual processing in unique ways, yet most models of vision focus on rigid bodies. Here, the authors use computational modeling to reveal that simulation-based intuitive physics underlies human soft object perception.

Extracellular matrix (ECM) remodeling is a hallmark of fibrosis thought to be driven by mesenchymal cells. Here, the authors discover that YAP-TEAD/LOX axis is activated in distal lung epithelial cells, which contributes to ECM remodeling in pre-clinical models of pulmonary fibrosis.

The authors develop multimodal machine learning models to infer metastatic recurrence risk for early-stage, hormone receptor-positive breast cancer from H&E images using >6000 cases across three centers, outperforming a nomogram and unimodal methods.