Search

Hund’s exchange interaction energy in two-dimensional materials is challenging to extract from experiments. Now, this is achieved in rhombohedral graphene, which allows an estimate of the interactions that drive the variety of correlated states in this material.

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.

Vallentgoed et al. integrate clinical and multiomic data from persons with matched initial and recurrent IDH-mutant astrocytomas to identify progression-associated mechanisms and report a DNA methylation-based signature associated with survival.

Treatment strategy of solid tumors requires continuous development. Here the authors develop a Macrophage Drug Conjugate (MDC) platform by loading ferritin-drug complexes to macrophages. MDC transfers the ferritin to cancer cells via ‘TRAnsfer of Iron binding protein’ (TRAIN) process and reduces tumor volume in various mouse tumor models.

The Faraday effect rotates the polarization plane of light in magneto-optical materials and is used for optical isolators blocking unwanted backscattering of light. Usually a small effect, Chin et al. have observed a large enhancement of the optical rotation by magneto-plasmonics.

In vertebrates parasite-mediated selection is thought to maintain polymorphism in MHC genes where specific resistance MHC alleles increase under emerging selection. Here, experimental evidence is shown from six stickleback fish populations that varying parasite selection helps maintain MHC polymorphism.

A single-cell ex vivo screening of repurposable drugs in glioblastoma and machine learning of drug–target networks show that anti-tumor neuroactive drugs converge on the AP-1/BTG pathway, based on which prediction models and experimental in vivo and in silico validation identify the anti-depressant vortioxetine as a potential therapeutic agent.

Although polymers are widely used in electric and photonic devices, the mechanism of charge transport across polymer chains is still not well understood. Here de Leeuw and colleagues propose a model that unifies experimental observations at high carrier densities.

Through archaeological excavation, morphological and proteomic taxonomic identification, mitochondrial DNA analysis and direct radiocarbon dating of human remains, a study reports the presence of Homo sapiens in Germany north of the Alps more than 45,000 years ago.

A neural epigenetic signature detectable via plasma analyses is prognostic in patients with glioblastoma, resembling an oligodendrocyte-progenitor- and neuronal-progenitor-cell-like state and showing increased neuro-to-glioma synapse formation.

Bloodstream infections with Candida parapsilosis in allogeneic hematopoietic cell transplantation recipients are associated with heteroresistance to the antifungal micafungin, which can be predicted from genomic features using machine learning.

Medulloblastoma in children is a difficult cancer to treat and the immune response to this tumour is not fully understood. Here the authors characterise and validate T cell epitopes from these cancers using an immunopeptidomics approach, comparing different molecular subtypes.

EGFRvIII-targeted CAR T cells have been proposed as a therapeutic option for patients with glioblastoma (GBM), however, clinical responses remain suboptimal. Here the authors engineer anti-EGFRvIII CAR T cells to secrete an optimized SIRPγ-derived CD47 blocker, showing that combining CAR T cell effector functions with enhanced macrophage-mediated tumor cell phagocytosis improves anti-tumor efficacy in preclinical models.

A global network of researchers was formed to investigate the role of human genetics in SARS-CoV-2 infection and COVID-19 severity; this paper reports 13 genome-wide significant loci and potentially actionable mechanisms in response to infection.

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.

The authors present palaeoclimatic data in the form of stable isotope records from equid teeth spanning 12,500 years of human occupation at the site of Ilsenhöhle in Ranis, Germany, including the earliest occupation of the site by Homo sapiens ~45,000 years ago.

Drawing microscopic information out of the diffusive dynamics of complex processes often requires an assumption of ergodicity. Precision experiments on a single atom in a periodic potential suggest that this may be too simplistic in many cases.

The ketone body β-hydroxybutyrate can be used as an alternative carbon source by T cells to maintain their function during severe respiratory viral infections, including infection with SARS-CoV-2.

A combination of genomic and clinical features improves predictions of response to immune checkpoint blockade.

Whole-genome sequencing, transcriptome-wide association and fine-mapping analyses in over 7,000 individuals with critical COVID-19 are used to identify 16 independent variants that are associated with severe illness in COVID-19.

The tumour microenvironment counteracts immune therapy in Glioblastomas. Authors show here, using spatially resolved and single cell transcriptomics, that dysfunctional T cells are induced by a myeloid cell subset via Interleukin-10 signalling, and inhibition of the downstream JAK/STAT pathway might restore glioblastoma immune therapy responsiveness.

Richard Lifton and colleagues identify a recurrent activating mutation in PRKACA, which encodes the catalytic subunit of protein kinase A, in cortisol-producing adrenal tumors. They further show that the mutation results in loss of binding by the regulatory subunit PRKAR1A, leading to increased phosphorylation of downstream targets.

Glioblastoma is divided into four subtypes based on molecular profiling at the methylome and transcriptome level. Here the authors perform an integrative analysis of these subtypes resulting in the identification of SOX10 whose loss induces a mesenchymal phenotype and promotes tumour progression.

Deep learning holds a great promise for the discovery and design of bioactive peptides, but experimental approaches to validate candidates in high throughput and at low cost are needed. Here, the authors combine deep learning and cell free biosynthesis for antimicrobial peptide (AMP) development and identify 30 functional AMPs, of which six with broad-spectrum activity against drug-resistant pathogens.

So far only a few compounds have been reported as splicing modulators. Here, the authors combine high-throughput screening, chemical synthesis, NMR, X-ray crystallography with functional studies and develop phenothiazines as inhibitors for the U2AF Homology Motif (UHM) domains of proteins that regulate splicing and show that they inhibit early spliceosome assembly on pre-mRNA substrates in vitro.

A chemical probe BI-9321 for the PWWP1 domain of NSD3 and its inactive analog were identified. BI-9321 binds to the methyl-lysine binding site, reduces the association of NSD3 with chromatin and inhibits proliferation of acute myeloid leukemia cells.

In this Review, Dai, Stockwell, Kroemer, Tang and colleagues offer a comprehensive discussion of the molecular regulation of ferroptosis and highlight how this may be potentially leveraged for therapeutic benefit for disease treatment.

Burying beetles feed their offspring on the carrion of vertebrate animals. Here, the authors study gene expression in the insect’s gut, as well as the composition of the microbiota in the gut and in carcasses, providing evidence for metabolic cooperation between host and specific microbes.

A recent report in Nature Cancer proposes that glioblastoma can be stratified into four functional subgroups on the basis of pathway inference from single-cell or bulk gene expression data. The analysis disclosed a pathway-based subgroup of glioblastoma that depends on oxidative phosphorylation and could potentially be targeted with mitochondrial metabolism inhibitors.

Current limitations of spintronics devices based on bulk topological materials stimulate the search for new materials and structures with interesting spin properties. Here the authors report a chiral spin texture around the Fermi level related to structural symmetry breaking in a TaSe₂ layer grown on a Bi2Se3 surface.

Common polygenic variation at chromosome 16p and rare recurrent deletions of 16p11.2 influencing autism risk are associated with reduced expression of genes throughout the 16p region, suggesting functional convergence of rare and common variant effects.

Jacobsen, Sherr et al. evaluate the utility of novel technologies in the treatment of type 1 diabetes. Their systematic review finds technologies such as continuous glucose monitoring, insulin pumps, and decision support tools improve important measures (e.g., HbA1c, time in range, quality of life) allowing precision-directed uptake of technology.

Here the authors provide a single-cell characterization of cerebrospinal fluid and blood of newly diagnosed multiple sclerosis (MS) patients, revealing altered composition of lymphocyte and monocyte subsets, validated by other methods including the interrogation of the TFH subset in mouse models of MS.

Some adolescents seek novelty, but it is unknown whether the brain circuits underlying this behaviour can be used to predict later, problematic behaviour. Here, authors show that diminished ventral striatal and prefrontal activity in response to anticipated rewards at age 14 in these individuals predicts problematic drug use at age 16.

A multi-laboratory study finds that single-molecule FRET is a reproducible and reliable approach for determining accurate distances in dye-labeled DNA duplexes.

Leptomeningeal metastatic disease (LMD) arises secondary to the metastatic dissemination of cancer cells into the leptomeninges and cerebrospinal fluid. Novel therapies against systemic disease have not yet translated into improved outcomes for patients with LMD, in whom median survival after diagnosis remains at 2–6 months. The authors of this Review, a multidisciplinary group of experts, describe the emerging evidence and areas of active investigation in LMD and provide directed recommendations for future research.

We introduce a major scientific endeavour called π-HuB (proteomic navigator of the human body), its aim being to generate and harness multimodality proteomic datasets to enhance our understanding of human biology.

Better analytical methods are needed to extract biological meaning from genome-wide association studies (GWAS) of psychiatric disorders. Here the authors take GWAS data from over 60,000 subjects, including patients with schizophrenia, bipolar disorder and major depression, and identify common etiological pathways shared amongst them.