Search

The BabyScreen+ study offered genomic screening to 1,000 newborns in Australia, and showed that the approach is feasible and positively received by families, leading to molecular diagnoses in 1.6% of babies.

Batatia and colleagues introduce a computational framework that combines message-passing networks with the atomic cluster expansion architecture and incorporates a many-body description of the geometry of molecular structures. The resulting models are interpretable and accurate.

Combining mobile tracking data and a survey experiment, Munzert et al. show that Germany’s contact tracing app is underused by those who socially distance less; however, even small cash incentives increased app uptake in the cohort.

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.

The paper presents a method that allows scaling machine learning interatomic potentials to extremely large systems, while at the same time retaining the remarkable accuracy and learning efficiency of deep equivariant models. This is obtained with an E(3)- equivariant neural network architecture that combines the high accuracy of equivariant neural networks with the scalability of local methods.

Individuals seldom reach out to old friends with whom they have lost touch. Interventions focused on changing attitudes were ineffective, but practicing reaching out to current friends first successfully encouraged people to reach out to old friends.

A protocol using large-scale training of graph networks enables high-throughput discovery of novel stable structures and led to the identification of 2.2 million crystal structures, of which 381,000 are newly discovered stable materials.

Ciranka, Linde-Domingo et al. show that inference of transitive orderings from pairwise relations benefits from a seemingly biased learning strategy, where observers update their belief about one of the pair members but not the other.

Noncoding transcripts contribute to the adaptation of cellular processes to oxygen levels. Here the authors characterize a hypoxia responsive lncRNA lincNORS and show that it has a role in cellular sterol homeostasis.

PENSIEVE-AI is a drawing-based, digital cognitive test that can be self-administered in <5 min. It matches traditional tests in detecting cognitive impairment and dementia, offering promise for early detection in literacy-diverse populations.

In this study, Rödström et al. use cryo-EM to reveal a range of structural features in THIK-1 K⁺ channel and demonstrate how these features define many functional properties of the channel, thus expanding its potential as a drug target.

An analysis of the impact of logging intensity on biodiversity in tropical forests in Sabah, Malaysia, identifies a threshold of tree biomass removal below which logged forests still have conservation value.

How accurate are social scientists in predicting societal change, and what processes underlie their predictions? Grossmann et al. report the findings of two forecasting tournaments. Social scientists’ forecasts were on average no more accurate than those of simple statistical models.

The authors introduce a diabatic neural network to accelerate excitedstate, non-adiabatic simulations of azobenzene derivatives. The model predicts quantum yields for unseen species that are correlated with experiment.

Complex experimental protocols often require multi-modal data acquisition with precisely aligned timing, as well as behaviour-dependent interventions. Here, authors describe a software with an intuitive user interface for flexible data acquisition without dedicated synchronisation hardware.

The sympathetic nervous system has been shown to respond to stress and activate haematopoietic stem cells. Here they show that cholinergic signals in the bone marrow preserve haematopoietic stem cell quiescence and self-renewal under proliferative stress.

Electric vehicles are only as green as the electricity used to charge them, but owners tend to charge vehicles at times of peak use. This study shows that tailored emails increase engagement with information about time-of-use tariffs, with maximal effects within the first three months of ownership.

The study provides a comprehensive transcriptomic atlas of the human gastrointestinal tract across the lifespan, highlighting inflammation-induced changes in epithelial stem cells that alter mucosal architecture and promote further inflammation.

An E(3)-equivariant deep learning interatomic potential is introduced for accelerating molecular dynamics simulations. The method obtains state-of-the-art accuracy, can faithfully describe dynamics of complex systems with remarkable sample efficiency.

The Indus river basin in South Asia is water stressed, energy insecure and intensively farmed, and research on this region often lacks a systemic approach to the issues. This study shows how the path to development in the region could be made less costly and more environmentally friendly by fostering transboundary cooperation.

The scale and nature of energy investments under diverging technology and policy futures is of great importance to decision makers. Here, a multi-model study projects investment needs under countries’ nationally determined contributions and in pathways consistent with achieving the 2 °C and 1.5 °C targets as well as certain SDGs.

Mutations in CTNS, the lysosomal cystine-proton symporter, cause cystinosis. Here authors report crystal structures of CTNS from Arabidopsis thaliana in complex with cystine, and establish the mode of ligand recognition and mechanism for proton-coupled cystine export from the lysosome.

Exosomes function as intercellular information transmitters and are candidates for delivery of therapeutic agents. Here the authors present EXOtic, a synthetic biology device for in-situ production of designer exosomes and demonstrate in vivo application in models of Parkinson's disease.

Achieving sustainable development goals while meeting the 1.5 °C climate target requires radical changes to how we use energy. A scenario of low energy demand shows how this can be done by down-sizing the global energy system to enable feasible deployment rates of renewable energy resources.

Opioids modulate pain, anxiety and stress by activating four subtypes of opioid receptors. The authors show that atypical chemokine receptor 3 (ACKR3) is a scavenger for various endogenous opioid peptides regulating their availability without activating downstream signaling.

Statistical mapping techniques provide insights into the spread of two key arbovirus vectors in Europe and the United States, and predict the future distributions of both mosquitoes in response to accelerating urbanization, connectivity and climate change.

Deep learning methods in natural language processing generally become more effective with larger datasets and bigger networks. But it is not evident whether the same is true for more specialized domains such as cheminformatics. Frey and colleagues provide empirical explorations of chemistry models and find that neural-scaling laws hold true even for the largest tested models and datasets.

This Registered Report presents evidence from 87 countries and regions showing that brief emotion-regulation interventions consistently reduced negative emotions and increased positive emotions during the COVID-19 pandemic.

To promote the recovery of the currently declining global trends in terrestrial biodiversity, increases in both the extent of land under conservation management and the sustainability of the global food system from farm to fork are required.

The predominant threats to pollinators vary across locations, as do perceptions of the consequences of pollinator loss. Here, the authors use formal expert elicitation methods to identify how pollination conservation experts rank the various drivers of pollinator decline and the range of risks to humans if pollination activity is lost.

McGovern and co-workers combine human brain signal measurements underlying decision formation with computational modelling to probe age-related differences in perceptual decision-making.

Superconducting quantum processors need to balance intentional disorder (to protect qubits) and nonlinear resonator coupling (to manipulate qubits), while avoiding chaotic instabilities. Berke et al. use the techniques of many-body localization theory to study the stability of current platforms against quantum chaos.