Computer vision for image-based transcriptomics

Abstract

Single-cell transcriptomics has recently emerged as one of the most promising tools for understanding the diversity of the transcriptome among single cells. Image-based transcriptomics is unique compared to other methods as it does not require conversion of RNA to cDNA prior to signal amplification and transcript quantification. Thus, its efficiency in transcript detection is unmatched by other methods. In addition, image-based transcriptomics allows the study of the spatial organization of the transcriptome in single cells at single-molecule, and, when combined with superresolution microscopy, nanometer resolution. However, in order to unlock the full power of image-based transcriptomics, robust computer vision of single molecules and cells is required. Here, we shortly discuss the setup of the experimental pipeline for image-based transcriptomics, and then describe in detail the algorithms that we developed to extract, at high-throughput, robust multivariate feature sets of transcript molecule abundance, localization and patterning in tens of thousands of single cells across the transcriptome. These computer vision algorithms and pipelines can be downloaded from: https://github.com/pelkmanslab/ImageBasedTranscriptomics. AbstractSingle-cell transcriptomics has recently emerged as one of the most promising tools for

Large‐scale image‐based profiling of single‐cell phenotypes in arrayed CRISPR‐Cas9 gene perturbation screens

“…We employed CellClassifier (https://www.pelkmanslab.org/?page_id=63) for data clean up and classification of transfected cells and cells in S‐phase of the cell cycle. We excluded missegmented cells, mitotic cells and cells displaying staining artefacts from further analysis (Stoeger et al , ). Computations were performed on the Brutus computing cluster (ETH Zürich) using the task manager iBRAIN.…”

Multivariate Control of Transcript to Protein Variability in Single Mammalian Cells

“…Additionally, we used bDNA sm-FISH with indirect immunofluorescence to quantify protein levels of 5 untagged genes ( Figure 1A). Using a previously established platform for image-based transcriptomics Stoeger et al, 2015), we obtained reproducible transcript counts in thousands of single cells, together with their relative background-corrected protein quantities derived from the integrated fluorescence intensities of the mEGFP or fluorescent antibody signal (Figures 1A and S1B-S1G; Tables S1-S3). We also inferred the cell-cycle position of cells by pulsing them with EdU for 15 min prior to the fixation (Gut et al, 2015).…”

Next-Generation Phenotypic Screening in Early Drug Discovery for Infectious Diseases

“…Interestingly, among a variety of optical methods aiming to overcome this major constraint, multibeam interferometric illumination [83] has yielded a commercialized cell-imaging system (Optical Biosystems, Santa Clara, CA, USA) capable of yielding optimized high-resolution equivalency in thousands of cells simultaneously using low-magnification optics. It will be interesting to see how such systems perform on highly multiplexed labeling of protein networks [84], and the transcriptome [85] visualized at mesoscopic scales. Such approaches could allow high-fidelity interactome analysis of host-cell/pathogen specific responses obviating background signal from noninfected cells.…”