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.

Federated ML (FL) provides an alternative to train accurate and generalizable ML models, by only sharing numerical model updates. Here, the authors present the largest FL study to-date to generate an automatic tumor boundary detector for glioblastoma.

A pangenome analysis of 76 wild and domesticated barley accessions in combination with short-read sequence data of 1,315 barley genotypes indicates that allelic diversity at structurally complex loci may have helped crop plants to adapt to agricultural ecosystems.

Since the introduction of transcatheter aortic valve implantation (TAVI) as a treatment for inoperable patients with aortic stenosis >1 decade ago, the range of patients who can benefit from this technology has continued to grow. In this Review, Jones and colleagues discuss the current evidence for TAVI in different patient populations, and describe the differences in design features in currently available valve systems.

Traditional type 2 diabetes (T2D) screening methods often depend on age, BMI guidelines and glucose measurements. Here, authors use a deep learning model that leverages chest radiographs and electronic health record data to screen for T2D, highlighting potential for early detection and intervention.

Previously, OLIG2-expressing stems cells have been identified as having a role in medulloblastoma recurrence. Here, the authors investigate the effects of targeting this OLIG2+ stem cell population in Sonic Hedgehog (SHH) medulloblastoma using CT-179, an OLIG2 inhibitor.

An extracorporeal cross-circulation approach enables, in a swine model, 36 hours of normothermic perfusion in healthy lungs, the recovery of injured lungs, and extended therapeutic interventions in all lungs.

By better quantifying uncertainties for marine ice-cliff instability, future Antarctic ice loss is predicted to be much lower than previously estimated.

The role of IgG glycosylation in the immune response has been studied, but less is known about IgM glycosylation. Here the authors characterize glycosylation of SARS-CoV-2 spike specific IgM and show that it correlates with COVID-19 severity and affects complement deposition.

Authors study antibody responses elicited against Omicron BA.1 and variants, by Omicron BA.1-bivalent mRNA-1273.214 vaccine compared with mRNA-1273 vaccination, when administered as a second booster, 90 days post-administration.

An Omicron BA.4/BA.5 mRNA booster vaccine elicits high neutralizing responses to the BA.4/BA.5 variant and to ancestral SARS-CoV-2, supporting tailoring booster vaccines to the predominant Omicron variant.

Solid organ transplant recipients are at increased risk of infectious disease and have unique molecular pathophysiology. Here the authors use host-microbe profiling to assess SARS-CoV-2 infection and immunity in solid organ transplant recipients, showing enhanced viral abundance, impaired clearance, and increased expression of innate immunity genes.

A bivalent vaccine encoding the ancestral and Beta variant SARS-CoV-2 spike proteins boosts neutralizing antibody responses against multiple viral variants, suggesting that a bivalent approach is an effective strategy to broadly enhance protection against COVID-19.

Although progress in the coverage of routine measles vaccination in children in low- and middle-income countries was made during 2000–2019, many countries remain far from the goal of 80% coverage in all districts by 2019.

Use of a comparative metagenomic approach suggests that phage communities in microbialites are endemic, and that marine and freshwater systems may share a common ancestral viral community.

Lifespan can be affected by both physiological ageing and specific sets of pathologies associated with old age. Here the authors report a resource of large-scale cross-sectional phenotyping of aging male mice at different time points to analyse a large set of phenotypes and molecular markers, including during genetic and diet interventions affecting lifespan.

In the first SARS-CoV-2 challenge study in humans, 36 young, healthy adult participants were intranasally inoculated with virus and monitored for productive infection, symptoms, virus kinetics, antibody response and safety in a controlled setting.

Genomic analyses of major clades of huge phages sampled from across Earth’s ecosystems show that they have diverse genetic inventories, including a variety of CRISPR–Cas systems and translation-relevant genes.

Post-acute sequelae of SARS-CoV-2 (PASC) is still not well understood. Here the authors provide patient reported outcomes from 590 hospitalized COVID-19 patients and show association of PASC with higher respiratory SARS-CoV-2 load and circulating antibody titers, and in some an elevation in circulating fibroblast growth factor 21.

Piccione et al find evidence for Antarctic ice sheet instability driven by millennial cycles in Southern Ocean temperature, providing clues for the mechanisms that link climate change and rapid Antarctic ice loss events.

Phenotypic variation and diseases are influenced by factors such as genetic variants and gene expression. Here, Barbeira et al. develop S-PrediXcan to compute PrediXcan results using summary data, and investigate the effects of gene expression variation on human phenotypes in 44 GTEx tissues and >100 phenotypes.

The authors summarize the data produced by phase III of the Encyclopedia of DNA Elements (ENCODE) project, a resource for better understanding of the human and mouse genomes.

This genome-wide study of age at first sexual intercourse and first birth identifies 371 signals driven by reproductive biology, externalising behaviour and environmental effects, with later onset associated with lower incidence of some diseases.

Alzheimer’s disease has been associated with increased structural brain aging. Here the authors describe a model that predicts brain aging from resting state functional connectivity data, and demonstrate this is accelerated in individuals with pre-clinical familial Alzheimer’s disease.

By comparing neural responses to diverse visual stimuli measured with a standardized two-photon imaging pipeline, the authors reveal response specializations within the mouse visual cortex.

Nearly 17 million sequences were used in a comparative metagenomic approach of 45 microbiomes and viriomes from distinct environments. Although core metabolic traits were represented in microbial genomes from all environments, differences could be correlated with biogeochemical condition. The analysis also revealed a surprising array of viral metabolic capabilities, suggesting that viruses may serve as a gene pool.

The Innovative Medicines Initiative Consortium RESOLUTE has started to develop tools and produce data sets to de-orphanize transporters in the solute carrier protein (SLC) superfamily, thereby lowering the barrier for the scientific community to explore SLCs as an attractive drug target class.

One hundred and ten Zika virus genomes from ten countries and territories involved in the Zika virus epidemic reveal rapid expansion of the epidemic within Brazil and multiple introductions to other regions.

Cataloging microbial genomes from Earth’s environments expands the known phylogenetic diversity of bacteria and archaea.

Mouse behavioral assays sheds light on the impact of adolescent exposure to vaping and the unique changes that occur in brain regions related to reward and reinforcement.

The increasing complexity of the implementation and operation of deep learning techniques hinders their reproducibility and deployment at scale, especially in healthcare. Pati and colleagues introduce a deep learning framework to analyse healthcare data without requiring extensive computational experience, facilitating the integration of artificial intelligence in clinical workflows.

The authors summarize the history of the ENCODE Project, the achievements of ENCODE 1 and ENCODE 2, and how the new data generated and analysed in ENCODE 3 complement the previous phases.