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.
An in silico approach helps researchers identify biosynthetic gene clusters coding for bioactive small molecules from metagenomic data of the human microbiome.
In two novel RNA modification mapping methods, the authors have engineered RNA enzymes and used the enzyme-mediated mutational signatures to map m6A and m1A at single-nucleotide resolution in mammalian RNA.
This Perspective highlights recent applications of deep learning in fluorescence microscopy image reconstruction and discusses future directions and limitations of these approaches.
The 2018 Data Science Bowl challenged competitors to develop an accurate tool for segmenting stained nuclei from diverse light microscopy images. The winners deployed innovative deep-learning strategies to realize configuration-free segmentation.
The 2018 Human Protein Atlas Image Classification competition sought to improve automated classification of protein subcellular localizations from fluorescence images. The winning strategies involved innovative deep learning approaches for multi-label classification.
Seamless integration of single-molecule localization microscopy and STED allows for correlative live imaging of protein position and movement at the nanoscale in the context of fine morphological features.
Tissue fixation with formaldehyde and a water-soluble carbodiimide crosslinker (EDC) leads to retention of extracellular vesicles within tissues and allows for reliable extracellular vesicle imaging for semiquantitative imaging applications.
Getting around the limitations of antibody-based N6-methyladenosine (m6A) pulldown, such as high input requirements and cross-reactivity, DART-seq profiles transcriptome-wide m6A occurrences from RNA amounts equivalent to the RNA obtained from 1,000 cells.
Robust m1A mapping of the human transcriptome is enabled by directed evolution of an HIV-1 reverse transcriptase with efficient read-through and high mutation rates at m1A sites along with development of new tools for data analysis.
Harmony, for the integration of single-cell transcriptomic data, identifies broad and fine-grained populations, scales to large datasets, and can integrate sequencing- and imaging-based data.
Directly sequencing RNA strands through a nanopore retains the full length of the transcript and allows for analysis of polyA tail length, transcript haplotypes and base modifications.
mmvec, a neural-network-based algorithm, uses paired multiomics data (microbial sequence counts and metabolite abundances) to compute the conditional probability of observing a metabolite in the presence of a specific microorganism.
UniRep learns fundamental protein features from millions of amino-acid sequences using a recurrent neural network. This summary of features can then be used for protein engineering.
Deep-Z uses deep learning to go from a two-dimensional snapshot to three-dimensional fluorescence images. The method improves imaging speed while reducing light dose, and was shown to be useful for accurate structural and functional imaging of neurons in Caenorhabditis elegans.