Articles

Improved green cAMP and red calcium sensors were developed to facilitate dual-color imaging in vivo. These sensors will allow studying the relationship between calcium and cAMP signaling.

scPROTEIN is a deep graph contrastive learning framework that can estimate the uncertainty of peptide quantification, denoise protein data, remove batch effects and encode single-cell proteomic-specific embeddings under a unified framework.

This work introduces two polishers for refining the draft genome generated from nanopore long reads, as well as an assembler pipeline for producing telomere-to-telomere diploid genome with low error rate.

TomoDRGN is a deep learning framework designed to model conformational and compositional heterogeneity from cryo-ET datasets on a per-particle basis.

An optogenetic system enables the controlled release of soluble and transmembrane proteins for precise exploration of cellular protein function at the single-molecule level and streamlined single-molecule imaging.

Implementation of ultralong transients on an Orbitrap mass spectrometer improves mass resolution, sensitivity and accuracy of charge determination in the analysis of large macromolecular ions.

Deng et al. expand the toolbox of neurotransmitter sensors with high-sensitivity green and red genetically encoded serotonin sensors. These are suitable for in vivo applications, as demonstrated in a variety of applications in mice.

A copper(II)-functionalized Mycobacterium smegmatis porin A nanopore enables direct identification of all 20 proteinogenic amino acids, one unnatural amino acid and two post-translational modifications, and shows potential for peptide discrimination and sequencing.

Effortless landmark detection is an unsupervised deep learning-based approach that addresses key challenges in landmark detection and image registration for accurate performance across diverse tissue imaging datasets.

mBaoJin is a monomeric derivative of the bright and photostable green fluorescent protein StayGold. mBaoJin offers favorable photophysical properties for use in diverse protein tagging and subcellular labeling applications.

Pretrained using over 33 million single-cell RNA-sequencing profiles, scGPT is a foundation model facilitating a broad spectrum of downstream single-cell analysis tasks by transfer learning.

A CasRx-based screening platform for genome-scale long noncoding RNA transcriptome perturbation is described.

Multi-sheet RESOLFT combines the speed and optical sectioning of light-sheet fluorescence microscopy with reversibly photoswitchable fluorescent proteins to enable fast, volumetric super-resolution imaging in live cells.

A-SOiD is a computational platform for behavioral annotation whose training includes elements of supervised and unsupervised learning. The approach is demonstrated on mouse, macaque and human datasets.

Targeting coalescent analysis (TarCA) is a statistical method that quantifies the number of progenitor cells of a given population using single-cell phylogenetic data.

TREX introduces an RNA-centric tool for identifying proteins binding to specific RNA regions in living cells.

The authors describe stem cell-derived bone marrow organoids that accurately model the structural and functional properties of the human bone marrow niche.

SLIDE is a machine learning-based method for identification of significant latent factors across multiomics datasets.

A biomimetic antigen-presenting system uses hexapod heterostructures to probe T-cell recognition at single-cell resolution.

Vibrational painting (VIBRANT) is a high-content single-cell phenotypic profiling method using mid-infrared imaging with vibrational probes for metabolic activity, which offers high accuracy with minimal batch effects to capture cellular responses to perturbation.