Search

A randomized controlled trial evaluating the selection of a single blastocyst for transfer by deep learning did not demonstrate noninferiority in clinical pregnancy rates when compared to trained embryologists using standard morphology criteria.

Mammalian genomes are scattered with repetitive sequences, but their biology remains largely elusive. Here, the authors show that transcription can initiate from short tandem repetitive sequences, and that genetic variants linked to human diseases are preferentially found at repeats with high transcription initiation level.

The anterior cingulate cortex encodes affective pain behaviours modulated by opioids; targeting opioid-sensitive neurons through a new chemogenetic gene therapy replicates the analgesic effects of morphine, providing precise chronic pain relief without affecting sensory detection.

The combination of JWST and ALMA data here unravel the history of the gas content of a quiescent galaxy, which became quenched through an act of self-sabotage. Black-hole accretion feedback heated the galaxy’s surrounding material, preventing its accretion.

This study introduces a sediment-based method to reconstruct Antarctic fast-ice change during the late Holocene, revealing cyclic patterns linked to solar variability and offering insight into long-term cryosphere climate dynamics.

Identifying jets originating from heavy quarks plays a fundamental role in hadronic collider experiments. In this work, the ATLAS Collaboration describes and tests a transformer-based neural network architecture for jet flavour tagging based on low-level input and physics-inspired constraints.

The distinct architecture of the Escherichia coli membrane transporter LetA mediates lipid trafficking across the bacterial envelope in partnership with the tunnel-like complex LetB.

Performing multiple histological stains on a biopsy can be costly and time consuming. Here the authors present a method for the digital transformation of H&E stained tissue into special stains (e.g., PAS, Masson’s Trichrome and Jones silver stain), and demonstrate that it improves diagnoses over the use of H&E only.

The famous nebula Barnard 68 has been used as a giant cosmic-ray detector: cosmic-ray-excited vibrational H₂ emission has been observed by JWST, giving a direct measurement of the CR ionization rate.

Dopaminergic medication and deep brain stimulation exert a shared therapeutic modulation of cortex-subthalamic nucleus network activity in Parkinson’s disease, through a suppression of high beta synchrony mediated by the hyperdirect pathway.

Green subsidies (carrots) are now becoming a more politically acceptable climate policy option compared with corrective regulations (sticks). However, researcher show that carrots without quick and appropriate sticks will not be sufficient to reach the deep decarbonization goal in the long run.

Pseudovirus-based SARS-CoV-2 spike deep mutational scanning is used to measure how mutations across the spikes affect ACE2 binding, cell entry or escape from human sera, producing data that could enable better prediction of viral evolution.

Cell spatial organization in tissue provides essential insights into diseases. Here, the authors show Ceograph, a graph convolutional network, for the analysis of pathology images to predict patient outcomes, highlighting cellular markers to guide personalized treatments and enhance biological understanding.

The impact of tumour heterogeneity on metastatic potential in prostate cancer remains poorly understood. Here, the analysis of single nuclei RNA sequencing and whole-genome sequencing from samples from five patients suggests an interplay between clonal evolution and cellular plasticity driving metastatic seeding.

Understanding biodiversity loss and achieving global commitments requires effective biodiversity monitoring. This Roadmap outlines the necessary steps to achieve a transnational European Biodiversity Observation Network built around Essential Biodiversity Variables, combining targeted sensing methods, spatial design, data sharing, data integration and modelling workflows, and coordinated governance to deliver policy-ready insights.

The relative contribution of lipid catabolism on fasting-induced longevity was unknown. Authors showed lifespan extension from fasting depend on silencing lipid catabolism upon nutrient replenishment through phosphorylation of NHR-49 by KIN-19.

The authors demonstrate a magnetic compute-in-memory platform that performs convolution using domain wall motion, enabling highly energy- and area-efficient processing for future AI hardware.

CELSR1 is a core component of the tissue/planar cell polarity pathway. Here, the authors report a CELSR1-related disorder featuring brain malformations, neurodevelopmental delay and epilepsy, and use mice to explore the pathophysiological mechanisms.

It is unclear whether microbes can efficiently degrade biodegradable plastics in the extreme environmental conditions of the seafloor. Here, Omura et al. show that biodegradable plastics can be degraded by the action of microorganisms on the deep-sea floor, although with much less efficiency than in coastal settings.

Verhaege et al. identify a conserved fibroblast barrier population at the base of the choroid plexus that compartmentalizes brain–CSF interfaces and is disrupted by inflammation, revealing a new site of central nervous system immune entry and barrier regulation.

A dataset of coding variation, derived from exome sequencing of nearly one million individuals from a range of ancestries, provides insight into rare variants and could accelerate the discovery of disease-associated genes and advance precision medicine efforts.

A genome-wide association study of cardiac magnetic resonance-derived left ventricular mass index including 43,000 UK Biobank participants reveals 12 associations (11 novel), implicating genes involved in cardiac contractility and cardiomyopathy.

The Chinese Liver Cancer Atlas project depicts a panoramic genomic landscape of hepatocellular carcinoma, covering candidate coding and non-coding drivers, mutational signatures, extrachromosomal circular DNA, subclonal catastrophic events and detailed evolutionary history.

Recurring deep ocean oxygen minimum events occur in the Pacific Southern Ocean during the warm climate interval Marine Isotope Stage 11. They are linked with perturbed Antarctic Bottom Water formation caused by West Antarctic Ice Sheet retreat.

In the deep sedimentary biosphere, 80 °C has been proposed as an upper thermal barrier for life. Using a suite of radiotracer experiments, this study reports active methanogenic and sulfate-reducing microbial populations with high cell-specific metabolic rates in deeply buried marine sediments from the Nankai Trough subduction zone, which reach temperatures up to 120 °C.

Effective early detection and intervention strategies for lung cancer are urgently needed. The authors of this Review summarize the current state of lung cancer screening and provide future directions for optimal implementation, including biomarker development. They also discuss precancer interception strategies that could transform both lung cancer prevention and early intervention.

CandyCrunch is a deep learning-based tool for predicting glycan structures from tandem mass spectrometry data. The paper also introduces CandyCrumbs that automatically annotates fragment ions in higher-order tandem mass spectrometry spectra.

By combining satellite observations with ground-based data and expert validation, this analysis demonstrates considerable misestimation of grassland extent and thereby carbon stock estimates in previous global assessments based on remote sensing.

Genome-wide association analyses identify variants associated with thoracic aortic diameter. A polygenic score for ascending aortic diameter was associated with a diagnosis of thoracic aortic aneurysm in independent samples.

Generative large language models (for example, ChatGPT and Claude) were used to score Big-Five traits from open-ended thoughts and video diaries. Large language model scores aligned with self-reports and predicted behaviour, showing personality’s expression in everyday language.

Suppressed Dexter transfer is needed to achieve efficient and stable hyperfluorescence, but complex matrices must be involved. A molecular design strategy has been proposed where Dexter transfer can be substantially reduced by an encapsulated terminal emitter, leading to ‘matrix-free’ hyperfluorescence.

Detecting amyloid deposits in tissue with Congo red can be limited by several factors, which can potentially lead to false diagnoses. Here, the authors use virtual birefringence imaging and virtual Congo red staining using autofluorescence of label-free human tissue, highlighting amyloid deposits in a consistent manner.

Deep proteome sequencing achieves ~80% coverage of the human proteome.

Sinking of organic matter represents the dominant mechanism for sequestration of anthropogenic CO₂in the deep sea. Here, the authors report the presence of healthy photosynthetic cells in the deep dark ocean, implying the fast injection of fresh organic carbon at depth across the global oligotrophic ocean.

Respiration enhances cerebrospinal fluid flow through mechanical and autonomic pathways. Inhale length and diaphragm motion influence its displacement and net flow, identifying a modifiable, noninvasive mechanism relevant to brain homeostasis.

In this Article, Singh and coauthors put forth a new machine-learning approach to evaluate inclusion and exclusion criteria from psychiatry abstracts to automate systematic reviews.

This study reveals that the ability to maintain a stable body weight predicts longevity in mice. By tracking weight dynamics throughout life, the authors identify genetic loci linked to body weight, energy homeostasis and healthy aging.

Integrating computational analyses of T cell exhaustion and mitochondrial fitness atlases with in vivo CRISPR screens has identified KLHL6 as a dual-negative regulator of both exhaustion differentiation and mitochondrial dysfunction, highlighting its potential as a target to enhance anti-tumour immunity.

Deep learning can automate the interpretation of medical imaging tests. Here, the authors formally assess the interchangeability of deep learning algorithms with expert human measurements for interpreting echocardiographic studies.

The authors develop a deep learning approach that enables an efficient search of the input space to find the best stimuli for modeled neurons. When tested, these stimuli are most effective at driving their matching cells in the brain.

Alzheimer’s disease is heterogeneous in its neuroimaging and clinical phenotypes. Here the authors present a semi-supervised deep learning method, Smile-GAN, to show four neurodegenerative patterns and two progression pathways providing prognostic and clinical information.

Here the authors present NCATS-SM0225, a small molecule that inhibits ERAD and selectively kills cancer cells by binding VDACs, disrupting calcium homeostasis, and triggering the PERK-STIM1 pathway to degrade ERAD regulators.