Skip to contents

Getting started

spacedeconv.Rmd

spacedeconv

spacedeconv is a unified interface to 31 deconvolution tools with focus on spatial transcriptomics datasets. In total 17 second-generation deconvolution tools are included, enabling deconvolution of any cell types when single-cell reference data is available. Additionally 10 first-generation tools, which are focussing on deconvolution of immune cells, are available as well as 4 first-generation methods optimised for mouse data. These methods don’t require scRNA-seq data to perform deconvolution as they utilize precomputed and verified signatures.

Installation

There are two ways to install spacedeconv:

  • The minimal installation installs only the dependencies required for the basic functionalities. All deconvolution methods need to be installed on-demand.
  • The complete installtation installs all dependencies including all deconvolution methods. This may take a considerable time.

Since not all dependencies are on CRAN or Bioconductor, spacedeconv is available from GitHub only. We recommend installing trough the pak package manager:

# install the pak package manager
install.packages("pak")

# recommended installation, deconvolution tools are installed on-demand
pak::pkg_install("omnideconv/spacedeconv")

# full installation including all deconvolution tools
pak::pkg_install("omnideconv/spacedeconv", dependencies = TRUE)

Data requirements

spacedeconv offers convenient access to perform first- and second-generation deconvolution on spatial transcriptomics datasets. While deconvolution can be performed directly with first-generation methods, second-generation algorithms require an additional annotated single-cell reference.

You can load 10X Visium Data by providing the spaceranger output folder. It is further possible to run spacedeconv with manually created SpatialExperiments. See the SpatialExperiment Documentation for further details.

spe <- SpatialExperiment::read10xVisium("path_to_directory")

Workflow

The core modules of spacedeconv are used to build a reference signature and deconvolute spatial transcriptomics data. We further provide convenient functions for normalization and visualization. A full list of deconvolution algorithms can be accessed by spacedeconv::deconvolution_methods or in the FAQ.

1. Normalization

spacedeconv offers a function to normalize spatial expression data. The normalization is saved in an additional assay, so make sure to use the correct data during deconvolution by providing the desired assay with the parameters assay_sc and assay_sp.

spe <- spacedeconv::normalize(spe, method = "cpm")

# specify expression assay to use
signature <- spacedeconv::build_model(spe,
  method = "quantiseq",
  assay_sp = "cpm"
)

2. Building a reference Signature

All second-generation deconvolution tools require a cell-type specific reference signature which can be calculated with build_model() function. To build a signature annotated single-cell sequencing data is required.

signature <- spacedeconv::build_model(
  single_cell_object,
  cell_type_col = "celltype_major",
  method = "spotlight",
  assay_sc = "cpm"
)

3. Deconvolution

This function deconvolutes spatial transcriptomics dataset using one of spacedeconvs 31 deconvolution tools. Since first-generation deconvolution tools contain predefined expression signatures these tools can be used directly on the spatial data. You do not need to provide a signature for this case. spacedeconv returns a SpatialExperiment object annotated with deconvolution results.

# save the results to an annotated SpatialExperiment
result <- spacedeconv::deconvolute(
  spatial_object,
  signature,
  method = "spotlight"
)

# return deconvolution results in table form
result <- spacedeconv::deconvolute(
  spatial_object,
  signature,
  method = "spotlight",
  return_object = FALSE
)

4. Visualization

We offer 6 visualization function aiming different insight of the data. A full example of each visualization option is available in the visualization Vignette.

# sample does refer to the first column of ColData(spe)
# for cell_type input a celltype present in the deconvolution result
spacedeconv::plot_celltype(spe, cell_type = "spotlight_B.cells")

# umi count
spacedeconv::plot_umi_count(spe)