Volume 16 Issue 1, January 2019

A probe for a journey into bubbles and why it’s good to be both inventor and discoverer.

Miniaturized, head-mounted fluorescent microscopes give researchers a clear view of neuronal activity as animals freely explore and interact with their surroundings.

Developments in imaging tools are making it possible to record activity from both large neuronal populations and subcellular components in freely moving animals. Although these developments are enabling relationships between brain activity and complex behaviors to be explored, many challenges need to be overcome before the potential of the freely moving animal can be fully utilized.

One major challenge in neuroscience is to uncover how defined neural circuits in the brain encode, store, modify, and retrieve information. Meeting this challenge comprehensively requires tools capable of recording and manipulating the activity of intact neural networks in naturally behaving animals. Head-mounted miniature microscopes are emerging as a key tool to address this challenge. Here we discuss recent work leading to the miniaturization of neural imaging tools, the current state of the art in this field, and the importance and necessity of open-source options. We finish with a discussion on what the future may hold for miniature microscopy.

The development of systems combining rapid volumetric imaging with three-dimensional tracking has enabled the measurement of brain-wide dynamics in freely behaving animals such as worms, flies, and fish. These advances provide an exciting opportunity to understand the organization of neural circuits in the context of voluntary and natural behaviors. In this Comment, we highlight recent progress in this burgeoning area of research.

Machine learning approaches that include deep learning are moving beyond image classification to change the way images are made.

Genetically encoded neurotransmitter sensors will shed light on neuronal communication.

Tools are needed to study phase behaviors of membraneless compartments in living cells.

Engineered genome structure explores function.

The rich repertoire of T cell receptors is awaiting large-scale profiling.

Synthetic approaches help to dissect the complexity of microbial interactions.

New technologies bring single-cell proteomics closer to reality.

Improved luciferases and luciferins are poised to shine.

A machine learning model predicts the genotype of CRISPR–Cas9 gene editing products, thereby enabling precise, template-free correction of disease-associated mutations.

A macroscope mounted on the head of a freely moving transgenic rat can help shine light on complex behaviors.

Synthetic small molecules can form liquid-phase condensates and simultaneously sequester intracellular enzymes.

A new development in high-speed atomic force microscopy enables microsecond resolution.

A vesicle-based mass spectrometry method enables the analysis of protein complexes in native membranes.

ERCC 2.0, spike-in controls and other metrics in genomics and metagenomics.

Expansion microscopy allows super-resolution images of diverse samples to be acquired on conventional microscopes, thus democratizing super-resolution imaging. This Perspective reviews available methods and provides practical guidance for users.

kBET informs attempts at single-cell RNA-seq data integration by quantifying batch effects and determining how well batch regression and normalization approaches remove technical variation while preserving biological variability.

The length of the guide RNA for Cas12a-VPR determines whether a target gene is edited or activated and allows for multiplexed, combinatorial gene modifications.

MPE-seq uses pools of reverse-transcription primers to facilitate enrichment of splice isoforms and pre-mRNA splicing intermediates.

DART-seq alters droplet sequencing in a simple and flexible way to simultaneously profile the transcriptome and multiplexed targeted RNAs, such as viral transcripts and immunoglobulin chains, in single cells.

A deep-learning-based tool, DeepNovo-DIA, uses data-independent-acquisition mass spectrometry data to sequence peptides without using a database.

A user-friendly ImageJ plugin enables the application and training of U-Nets for deep-learning-based image segmentation, detection and classification tasks with minimal labeling requirements.

U-ExM enables near-native expansion microscopy of samples in vitro and in cells. The combination of U-ExM with confocal microscopy and HyVolution revealed details of centriole chirality that were previously accessible only by electron microscopy.

A protocol adapted to xeno- and feeder-free conditions is shown to generate reliable and consistent cortical brain organoids across differentiations and source stem cell lines, making it suitable for disease modeling and other applications.

A transcriptional analysis of kidney organoids reveals batch effects as the key drivers of variation, mainly through differences in maturity, and provides a list of highly variable genes and a method for estimating differentiation stage for improved disease modeling.

Segmental Duplication Assembler (SDA) uses long sequence reads to resolve segmental duplications that are collapsed in current genome assemblies. These assemblies correspond in total to the length of an average human chromosome.

The DNA-based, ratiometric fluorescent reporter CalipHluor enables quantitative imaging of pH and calcium in acidic organelles with single-organelle resolution. The probe was used to identify a lysosome-specific Ca²⁺ importer in animals.

Deep learning enables cross-modality super-resolution imaging, including confocal-to-STED and TIRF-to-TIRF-SIM image transformation. Imaging of a larger FOV and greater depth of field is possible with higher resolution and SNR at lower light doses.

fMRIPrep is a robust and easy-to-use pipeline for preprocessing of diverse fMRI data. The transparent workflow dispenses of manual intervention, thereby ensuring the reproducibility of the results.

LEAP is a deep-learning-based approach for the analysis of animal pose. LEAP’s graphical user interface facilitates training of the deep network. The authors illustrate the method by analyzing Drosophila and mouse behavior.

A compressed sensing approach enables the identification of key neurons involved in a particular behavior with few measurements, using genetic tools with limited specificity. The approach is demonstrated in the C. elegans interneuron circuitry.