Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript.
Minimal tumour perturbations that boost T-cell infiltration can be discovered by using deep learning to analyse large-scale spatial omics profiles of tumours.
A positron emission tomography tracer of reactive oxygen and nitrogen species enables the in vivo quantification of oxidative stress, as shown in mouse models of oxidative stress, stroke and tauopathy.
The pretraining efficiency and generalization of medical foundation models can be enhanced by using synthetic data generated via conditioning with disease labels.
Therapeutic antibodies with enhanced resistance to the evolution of SARS-CoV-2 can be identified via deep mutational learning by screening a combinatorial library of receptor-binding domains for antibody binding and escape.
3D-printed models of the penis can facilitate the study of penile physiology and the repair of penile defects, as shown via a reconstructed and implanted corpus cavernosum that restored erectile function and mating capacity in rabbits and pigs.
Radiation-therapy-induced DNA damage in oral and rectal epithelial cells can be reduced via the nanoparticle-mediated local delivery of messenger RNA encoding for a damage-suppressor protein found in radiation-resistant tardigrades.
The wrapping of single bacteriophage-infected bacteria with a polymeric nanoscale coating preserves the vitality of the replicating phages, which led to enhanced therapeutic outcomes in mice with bacterially induced enteritis and arthritis.
Self-supervised representation learning on data from imaging mass cytometry can be used to distinguish morphological differences in tumour microenvironments and to precisely characterize distinct microenvironment signatures.
Clinically licensed nitroglycerin patches allowed for the on-demand and sustained expression of glucagon-like peptide-1 by human cells engineered with a nitroglycerin-responsive gene switch subcutaneously implanted in mice with type 2 diabetes.
By mimicking antigen processing and presentation during the rejection of xenotransplanted tissue, xenogeneic cell membrane-derived vesicles encapsulating peptide antigens or mRNA-encoded antigens function as vaccines targeted to dendritic cells.
Mass cytometry can be used to scale up multiplexed serology testing, as shown for the simultaneous detection of antibody levels against multiple SARS-CoV-2 proteins in hundreds of serum samples.
A gut-on-a-chip incorporating peristaltic-like movements and faecal samples from patients with melanoma recapitulates microbiomeâhost interactions and predicts the responses of the patients to immune checkpoint inhibitors.
The parallel intracellular recording of neurons via an array of 4,096 microhole electrodes on a semiconductor chip enables the large-scale mapping of synaptic connections, as shown with rat neuronal cultures.
Small circular RNAs encoding one or more antigens and enclosed in subcutaneously injected lipid nanoparticles can elicit potent and durable antitumour T cell responses for robust tumour immunotherapy, particularly in combination with immune checkpoint inhibition, as shown in multiple mouse models of tumours.
Variants of adeno-associated viruses displaying robust and widespread transduction in murine, porcine and non-human-primate kidneys, and in human kidney organoids, can be obtained via the cross-species cycling of capsid libraries in the different kidney models.
Urinary reporters leveraging caged bioorthogonal click handles for the biomarker-dependent activation of âclickabilityâ and renal clearance allow for the monitoring, in urine, of the efficacy of chemotherapy and of associated kidney injury in mice.
An interpretable transformer-based model leveraging graph representation learning accurately predicts cancer genes across homogeneous and heterogeneous pan-cancer networks of biological interactions.
A skull-stripping model for magnetic resonance images that leverages personalized priors from atlases of brain scans generates naturally consistent and seamless lifespan changes in brain volume.
A prophylactic cancer vaccine consisting of commercial adjuvants wrapped by membranes from pluripotent stem cells inhibited tumour progression in mice, owing to shared antigens between tumour membranes and the pluripotent stem cell membranes.
Metabolic subphenotypes of type 2 diabetes can be predicted by the shape of the glucose curve measured via a continuous glucose monitor during standardized oral glucose-tolerance tests performed at home.
