SpatialOmicsOverlay: Overlay of Structurally Profiled Regions of Interest on GeoMx DSP Tissue Images
Maddy Griswold, with NanoString Technologies, gave a workshop at the BioConductor Conference 2022. Griswold’s workshop was on SpatialOmicsOverlay: Overlay of Structurally Profiled Regions of Interest on GeoMx DSP Tissue Images.
From the Bioconductor Conference 2022
Author(s): Maddy Griswold, David Allen Henderson, Nicole Ortogero
Affiliation(s): NanoString Technologies
Spatial biology is changing the way we view biology. We can now look at a tissue in terms of compartments such as cells, tissue structure, or disease state. Compartments are selected on a tissue image by a scientist or pathologist but then that image is not typically used after compartment selection. Data analysis and visualization then proceed as in a typical RNA-seq or scRNA-seq analysis, but this data and the story around it is amplified when presented in conjunction with the image. At present, analysis packages lack the ability to handle the free-hand Region of Interest (ROI) selection capability available in the NanoString GeoMx instrument workflow. Each ROI can be drawn in any shape imaginable and can be even further segmented to Areas of Interest within an ROI that follows the morphology of the tissue. The immense range of possible sizes and shapes across an image presents a challenge when trying to overlay data on the image. To address this issue, we created the SpatialOmicsOverlay package. SpatialOmicsOverlay was written for NanoString’s GeoMx and CosMx data but should work on any OME-TIFF file. From the OME-TIFF, the XML and image are extracted. The XML file contains all of the spatial information necessary to position the ROIs in the correct location on the image. With the addition of ROI annotations, users can view the compartments selected on the tissue image colored by their annotations of interest. These annotations can be almost anything from tissue structure, disease state, gene expression or pathway score. Users can easily manipulate their images and overlays: flipping axes, color correcting, and cropping. All plots are ggplot based which makes figure customization easy. Future releases are planned to include functionality to graph on top of the image and perform image analysis. The SpatialOmicsOverlay package fills a gap in current R packages for the burgeoning spatial biology field.
Package demo details
https://nanostring.com/wp-content/upl...
Source code
https://nanostring.com/resources/geos...
More Resources
Bioconductor Conference Site: https://bioc2022.bioconductor.org/
BioC2022 Github: https://github.com/Bioconductor/BioC2022
Main Site: https://www.r-consortium.org/
News: https://www.r-consortium.org/news
Twitter: / rconsortium
LinkedIn: / r-consortium