Articles

DeePiCt (deep picker in context) is a versatile, open-source deep-learning framework for supervised segmentation and localization of subcellular organelles and biomolecular complexes in cryo-electron tomography.

This work presents a computational framework, COMMOT, to spatially infer cell–cell communication from transcriptomics data based on a variant of optimal transport (OT).

An optimized pipeline for improved inference and analysis of structural variants (SVs) has been developed, which uses Iris for refining breakpoints and sequences, and Jasmine for comparing SV calls at population scale.

Hybrid Unmixing offers enhanced imaging of multiplexed fluorescence labels, enabling longitudinal imaging of multiple fluorescent signals with reduced illumination intensities.

A computational pipeline for haplotype-aware pantranscriptome analysis has been developed, which enables spliced pangenome graph construction, RNA sequencing data alignment, and estimation of haplotype-specific transcript expression levels.

Cryogenic correlated light, ion and electron microscopy (cryo-CLIEM) integrates three-dimensional confocal microscopy with focused ion beam–scanning electron microscopy for efficient preparation of lamellae containing target structures for in situ structural biology with cryo-electron tomography.

The ELI-TriScope advances cryo-CLEM by focusing light, electron and ion beams on cryopreserved samples for markedly improved preparation of cryo-lamellae containing target structures.

MARIO is a robust tool for accurately matching multimodal single-cell datasets.

The Photopick platform, which can be used for phenotype-activated cell selection, was used to develop the improved voltage sensors QuasAr6a and QuasAr6b. These GEVIs offer improved signals and are useful for all-optical electrophysiology.

This paper presents nonnegative spatial factorization, a general framework for spatially aware and interpretable dimension reduction for high-dimensional spatial data, and its application to spatial transcriptomics analysis.

A statistical approach for optimal design of multiplexed imaging studies has been developed. It determines experimental parameters that facilitate cell phenotype identification.

During segmentation of neurons in electron microscopy datasets, auxiliary learning via the prediction of local shape descriptors increases efficiency, which is important for the processing of datasets of ever-increasing size.

A physiologically relevant 3D model of the outer-blood-retina-barrier recapitulates healthy and disease phenotypes.

Virus-assisted directed evolution of tRNAs (VADER) enables selection of improved suppressor tRNAs in mammalian cells. Improved pyrrolysyl- and tyrosyl-tRNA variants enhance unnatural amino acid incorporation in mammalian systems.

SVDSS combines multiple strategies to improve detection of structural variation using low-error long-read sequencing data.

This work introduces CaRPool-seq that utilizes the RNA-targeting CRISPR–Cas13d system to perform combinatorial perturbations in single-cell screens.

Nano3P-seq presents a nanopore-based sequencing tool to profile polyA-tailed and non-polyA-tailed transcripts, as well as capture polyA tail length and composition.

Localization Model Fit (LocMoFit) is an open-source tool for extracting meaningful parameters from individual structures in localization microscopy data. The framework was used for quantitative analysis of diverse biological structures.

This paper shows that the uniformity of vitreous ice thickness relies on the surface flatness of the supporting film, and presents a method to use ultraflat graphene as the support for cryo-EM specimen preparation.

DEDAL is a deep learning-based protein sequence alignment method that improves the quality of predicted alignment for remote homologs and better discriminates remote homologs from evolutionarily unrelated sequences.