News & Views

Focused laser light can manipulate the phase of electrons and solve the long-standing phase contrast conundrum in electron microscopy.

A widely used concept from machine learning is put to use for single-cell analysis

Single objective light-sheet fluorescence microscopes combine the convenience of conventional sample mounting with sensitive subcellular and super-resolution imaging of cells and tissues.

Understanding of fundamental questions in human embryology is hampered by limited access to in utero developmental events. Manfrin et al. have engineered an in vitro platform that recapitulates early morphogenic events of human development with unprecedented spatial and temporal control.

The interpretation of fragmentation patterns in tandem mass spectrometry is crucial for peptide sequencing, but the relative intensities of these patterns are difficult to predict computationally. Two groups have applied deep neural networks to address this long-standing problem in the proteomics field, extending theoretical spectra with an additional dimension of high-accuracy fragment ion intensities.

Light-sheet fluorescence microscopes are making it possible to follow subcellular dynamics with unprecedented detail, but they are complex to build and maintain. A new method that can generate arbitrary light sheets provides much-needed simplicity and additional versatility.

A recent approach for single-cell RNA-sequencing data uses Bayesian deep learning to correct technical artifacts and enable accurate and multifaceted downstream analyses.

Two cell-based resources producing a broad repertoire of glycosaminoglycan structures will facilitate new applications in the glycosciences field and beyond.

A type of neural network first described in 2015 can be trained to translate between images of the same field of view acquired by different modalities. Trained networks can use information inherent in grayscale images of cells to predict fluorescent signals.

Building on methods from electron microscopy, researchers are applying nanoscale fluorescence imaging to address questions in structural biology.

A new article by Pandarinath et al. describes an artificial neural network model that captures some key aspects of the activity of populations of neurons in the primary motor cortex.

A new detector built for X-ray free-electron lasers provides unprecedented speed and accuracy for macromolecular crystallography at synchrotron radiation facilities—and finally allows crystallographers to harness the full capabilities of those sources.

Chemically modified DNA aptamers enable quantitative super-resolution imaging.

Quanti.us is a platform that allows large-scale manual image annotation by a distributed workforce through Amazon’s Mechanical Turk crowdsourcing marketplace.

Overexpression of mouse thymic-stromal-cell-derived lymphopoietin in immune-compromised mice that harbor a reconstituted human immune system rescues lymph node formation and enhances adaptive immune responses.

Two methods for controlling protein activities with small molecules provide a general solution to a long-standing challenge in mammalian synthetic biology.

Two methods for comparing single-cell expression data sets help address the challenge of integrating data across conditions and experiments.

An interpretable, deep neural network produces mechanistic hypotheses on how genetic interactions contribute to whole-cell phenotypes.

The combination of metabolic RNA labeling with biochemical nucleoside conversion now adds a broadly applicable temporal dimension to RNA sequencing.

Two complementary approaches improve 3D mass spectrometry imaging of biomolecules on 3D surfaces and within tissue sections and single cells.