Here we demonstrate the SpatialFeatureExperiment (SFE), an S4 class that brings Simple Features (as in the sf package) to SpatialExperiment, and exploratory spatial data analysis (ESDA) with the SFE object with the Voyager package. This workshop has 3 parts:

  1. An introduction to spatial transcriptomics, geospatial data, and how the geospatial methods may or may not apply to spatial transcriptomics.
  2. Introduction to the sf data frame and the SFE object, which bring geospatial tools to spatial transcriptomics. A published Visium dataset already as an SFE object provided in the SFEData package is used for the demonstrations here and in part 3.
  3. Plotting and basic ESDA with the Voyager package, applying geospatial tools on the SFE object. The ESDA includes computing Moran’s I and Geary’s C for spatial autocorrelation, permutation testing for Moran’s I, correlograms, and Moran scatter plot.

Prerequisites

  • Basic knowledge of R syntax and data structures
  • Familiarity with SingleCellExperiment and preferably also SpatialExperiment classes
  • Familiarity with 10X Visium
  • Familiarity with sf is preferable

Installation

We have submitted the packages presented in this workshop to Bioconductor. Before they get accepted, they can be installed from GitHub:

devtools::install_github("pachterlab/SFEData")
devtools::install_github("pachterlab/SpatialFeatureExperiment")
devtools::install_github("pachterlab/Voyager")

Using the Docker image

docker run -e PASSWORD=abc -p 8787:8787 ghcr.io/lambdamoses/sfeworkshop:latest

Once running, navigate to http://localhost:8787/ and then login with username rstudio and password abc.