Search

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.

Kumar et al. show that large language models (LLMs) nearly match expert reliability and outperform laypeople when assessing empathic communication across multiple frameworks. The performance of both LLMs and experts depends on clear and specific evaluation criteria.

Openshaw and colleagues find myeloid inflammation and complement activation signatures in patients with long COVID who were previously hospitalized.

The lowest-frequency gravitational wave background may be shaped by supermassive black hole binaries that scatter nearby stars or dark matter. In this case, the NANOGrav 15-year dataset favours dense galactic centres with 10⁶ solar masses per cubic parsec.

The Omicron variant evades vaccine-induced neutralization but also fails to form syncytia, shows reduced replication in human lung cells and preferentially uses a TMPRSS2-independent cell entry pathway, which may contribute to enhanced replication in cells of the upper airway. Altered fusion and cell entry characteristics are linked to distinct regions of the Omicron spike protein.

Pooling participant-level genetic data into a single analysis can result in variance stratification, reducing statistical performance. Here, the authors develop variant-specific inflation factors to assess variance stratification and apply this to pooled individual-level data from whole genome sequencing.

Here, the authors examine the mechanisms behind cheatgrass’s successful invasion of North American ecosystems. Their genetic analyses and common garden experiments demonstrate that multiple introductions and migrations facilitated cheatgrass local adaptation.

The goals, resources and design of the NHLBI Trans-Omics for Precision Medicine (TOPMed) programme are described, and analyses of rare variants detected in the first 53,831 samples provide insights into mutational processes and recent human evolutionary history.

Large genome-wide meta-analysis of clinically diagnosed late-onset Alzheimer’s disease (LOAD) from 94,437 individuals identifies new LOAD risk loci and implicates Aβ formation, tau protein binding, immune response and lipid metabolism.

Most active particles studied to date lack the ability to undergo controlled shape transformations and control over their propulsion in response to environmental stimuli. Here, the authors present a class of active particles made from stimuli-responsive materials that exhibit fully reversible shape-dependent propulsion.

Here the authors provide an explanation for 95% of examined predicted loss of function variants found in disease-associated haploinsufficient genes in the Genome Aggregation Database (gnomAD), underscoring the power of the presented analysis to minimize false assignments of disease risk.

Excised signal circles are circular DNA by-products of V(D)J recombination that form a complex with the V(D)J recombinase, and when increased in abundance, result in increased mutagenesis, causing adverse outcomes in cancer.

The potential roles of viruses in microbial-induced permafrost thaw, which is accelerated by climate change, are unclear. Here, Trubl et al. sample a permafrost thaw gradient and identify thousands of new viruses carrying genes with metabolic-related functions, which might influence the fate of stored carbon.

Chronic infection with SARS-CoV-2 leads to the emergence of viral variants that show reduced susceptibility to neutralizing antibodies in an immunosuppressed individual treated with convalescent plasma.

In this study, long-read RNA sequencing achieves accurate single-nucleotide polymorphism calling, haplotype phasing and allele-specific expression analysis.

H5N1 avian influenza viruses caused an outbreak in dairy cattle. We show that the potential for avian viruses to replicate in cow cells varies across H5N1 evolution, suggesting that the risk of spillover into mammals differs between variants.

A genomic constraint map for the human genome constructed using data from 76,156 human genomes from the Genome Aggregation Database shows that non-coding constrained regions are enriched for regulatory elements and variants associated with complex diseases and traits.

RMC-7977, a compound that exhibits potent inhibition of the active states of mutant and wild-type KRAS, NRAS and HRAS variants has a strong anti-tumour effect on RAS-addicted tumours and is well tolerated in preclinical models.

Sera from vaccinated individuals and some monoclonal antibodies show a modest reduction in neutralizing activity against the B.1.1.7 variant of SARS-CoV-2; but the E484K substitution leads to a considerable loss of neutralizing activity.

Gasdermin D (GSDMD) is a pore forming protein activated by inflammasome derived caspases. Here the authors characterize the function of GSDMD in mouse influenza virus infection and show that immunopathology is reduced in the absence of GSDMD and involves changes in neutrophil function.

Insights into regulatory T cell biology are accelerating therapeutic innovation in cancer immunotherapy, autoimmune diseases and transplant rejection.

The goal of the 1000 Genomes Project is to provide in-depth information on variation in human genome sequences. In the pilot phase reported here, different strategies for genome-wide sequencing, using high-throughput sequencing platforms, were developed and compared. The resulting data set includes more than 95% of the currently accessible variants found in any individual, and can be used to inform association and functional studies.

Here, the authors leverage whole-genome sequencing from >88,000 diverse individuals to uncover 18 BMI-related genetic signals, including novel variants in participants of African genetic ancestry, highlighting the value of diversity in genetic discovery.

Here, the authors develop pluripotent stem cell-derived kidney organoids by mimicking in vivo proximal differentiation, providing a model to study nephron development, injury responses, and a platform for therapeutic discovery.

Platelet aggregation is associated with myocardial infarction and stroke. Here, the authors have conducted a whole genome sequencing association study on platelet aggregation, discovering a locus in RGS18, where enhancer assays suggest an effect on activity of haematopoeitic lineage transcription factors.

This report from the 1000 Genomes Project describes the genomes of 1,092 individuals from 14 human populations, providing a resource for common and low-frequency variant analysis in individuals from diverse populations; hundreds of rare non-coding variants at conserved sites, such as motif-disrupting changes in transcription-factor-binding sites, can be found in each individual.

Glioblastoma (GBM) is characterized by a high degree of heterogeneity and plasticity due to interplay with neural developmental programs. Here, the authors develop a model of GBM by introducing sequential oncogenic mutations in human neural stem cells and using this, identify INSM1 as a driver of a neural progenitor gene network promoting tumorigenesis.

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.

In a trial involving 45 clinics and 15,596 visits, the use of an AI-based screening tool applied to ECG for opportunistic detection of advanced chronic liver disease led to a significant increase in new diagnoses compared to the standard workflow.

An approach combining BONCAT, stable isotope probing and metaproteomics showcases the hidden metabolic interconnectivity of microorganisms within an anaerobic digestion community.

Turley and colleagues find that Gremlin1-expressing lymph node fibroblastic reticular cells support dendritic cell homeostasis via provision of FLT3L.

In this study, Aggarwal and colleagues perform prospective sequencing of SARS-CoV-2 isolates derived from asymptomatic student screening and symptomatic testing of students and staff at the University of Cambridge. They identify important factors that contributed to within university transmission and onward spread into the wider community.

Using data from a single time point, passenger-approximated clonal expansion rate (PACER) estimates the fitness of common driver mutations that lead to clonal haematopoiesis and identifies TCL1A activation as a mediator of clonal expansion.

Active galactic nuclei are surrounded by a dusty and molecular disk that fuels supermassive black holes and connects them to their host galaxies. Here, the authors show with JWST interferometric observations that most of the dust in the Circinus galaxies lies in a compact disk, while only a tiny fraction traces hot outflowing material.

Genetic and testing data from England show that the SARS-CoV-2 variant of concern B.1.1.7 has a transmission advantage over other lineages.

STAAR is a powerful rare variant association test that incorporates variant functional categories and complementary functional annotations using a dynamic weighting scheme based on annotation principal components. STAAR accounts for population structure and relatedness and is scalable for analyzing large whole-genome sequencing studies.

Genomic analyses applied to 14 childhood- and adult-onset psychiatric disorders identifies five underlying genomic factors that explain the majority of the genetic variance of the individual disorders.

Post-international travel quarantine has been widely implemented to mitigate SARS-CoV-2 transmission, but the impacts of such policies are unclear. Here, the authors used linked genomic and contact tracing data to assess the impacts of a 14-day quarantine on return to England in summer 2020.

Large-effect variants in autism remain elusive. Here, the authors use long-read sequencing to assemble phased genomes for 189 individuals, identifying pathogenic variants in TBL1XR1, MECP2, and SYNGAP1, plus nine candidate structural variants missed by short-read methods.

Bosonic qubits can be engineered to feature intrinsic protection against certain kinds of errors, which makes quantum error correction across many bosonic qubits possible with less overhead.

A versatile, human iPSC-derived nephron engineering platform that permits scrutiny of axial patterning mechanisms is critical for identifying the origins of human kidney disease. Here they describe a system in which synchronized human nephron structures are generated from pluripotent stem cells, enabling manipulation of axial segmentation.

This study describes the integrative analysis of 111 reference human epigenomes, profiled for histone modification patterns, DNA accessibility, DNA methylation and RNA expression; the results annotate candidate regulatory elements in diverse tissues and cell types, their candidate regulators, and the set of human traits for which they show genetic variant enrichment, providing a resource for interpreting the molecular basis of human disease.

COVID-19 can be associated with neurological complications. Here the authors show that markers of brain injury, but not immune markers, are elevated in the blood of patients with COVID-19 both early and months after SARS-CoV-2 infection, particularly in those with brain dysfunction or neurological diagnoses.

A global dataset of the satellite-tracked movements of pelagic sharks and fishing fleets show that sharks—and, in particular, commercially important species—have limited spatial refuge from fishing effort.

The age of the Grand Canyon is fervently debated. Thermochronological reconstructions of canyon incision show that although parts of the canyon were carved more than 50 million years ago, two key segments formed less than 6 million years ago, implying that the canyon is a young feature.

A study shows that clonal haematopoiesis of indeterminate potential is associated with an increased risk of chronic liver disease specifically through the promotion of liver inflammation and injury.

Analysis of 97,691 high-coverage human blood DNA-derived whole-genome sequences enabled simultaneous identification of germline and somatic mutations that predispose individuals to clonal expansion of haematopoietic stem cells, indicating that both inherited and acquired mutations are linked to age-related cancers and coronary heart disease.

The Zika viral protease NS2B-NS3 is a crucial target for antiviral drug development due to its role in processing viral polyproteins. Here, the authors utilize crystallographic fragment screening and deep mutational scanning to identify binding sites for resistance-resilient inhibitors.

Using well-calibrated statistical methods the authors jointly analyze Undiagnosed Diseases Network genomes, identifying known and novel disease genes. Software is publicly available to support future cross-cohort rare disease discovery efforts.