In TrigosTeam/SPIAT: Spatial Image Analysis of Tissues

knitr::opts_chunk$set(echo = TRUE)

library(SPIAT)

In this vignette we will use an inForm data file that's already been formatted for SPIAT with format_image_to_spe(), which we can load with data(). We will use define_celltypes() to define the cells with certain combinations of markers.

data("simulated_image")

# define cell types
formatted_image <- define_celltypes(
    simulated_image, 
    categories = c("Tumour_marker","Immune_marker1,Immune_marker2", 
                   "Immune_marker1,Immune_marker3", 
                   "Immune_marker1,Immune_marker2,Immune_marker4", "OTHER"), 
    category_colname = "Phenotype", 
    names = c("Tumour", "Immune1", "Immune2", "Immune3", "Others"),
    new_colname = "Cell.Type")

We will be performing some basic analyses on this image. Here is the visualisation of the cell coordinates.

my_colors <- c("red", "blue", "darkcyan", "darkgreen")

plot_cell_categories(spe_object = formatted_image, 
                     categories_of_interest = 
                       c("Tumour", "Immune1", "Immune2", "Immune3"), 
                     colour_vector = my_colors, 
                     feature_colname = "Cell.Type")

Cell percentages

We can obtain the number and proportion of each cell type with calculate_cell_proportions(). We can use reference_celltypes to specify cell types to use as the reference. For example, "Total" will calculate the proportion of each cell type against all cells. We can exclude any cell types that are not of interest e.g. "Undefined" with celltypes_to_exclude.

p_cells <- calculate_cell_proportions(formatted_image, 
                                      reference_celltypes = NULL, 
                                      feature_colname ="Cell.Type",
                                      celltypes_to_exclude = "Others",
                                      plot.image = TRUE)
p_cells

Alternatively, we can also visualise cell type proportions as barplots using plot_cell_percentages().

plot_cell_percentages(cell_proportions = p_cells, 
                      cells_to_exclude = "Tumour", cellprop_colname="Proportion_name")

Cell distances

Pairwise cell distances

We can calculate the pairwise distances between two cell types (cell type A and cell type B) with calculate_pairwise_distances_between_cell_types(). This function calculates the distances of all cells of type A against all cells of type B.

This function returns a data frame that contains all the pairwise distances between each cell of cell type A and cell type B.

distances <- calculate_pairwise_distances_between_celltypes(
  spe_object = formatted_image, 
  cell_types_of_interest = c("Tumour", "Immune1", "Immune3"),
  feature_colname = "Cell.Type")

The pairwise distances can be visualised as a violin plot with plot_cell_distances_violin().

plot_cell_distances_violin(distances)

We can also calculate summary statistics for the distances between each combination of cell types, the mean, median, min, max and standard deviation, with calculate_summary_distances_between_celltypes().

summary_distances <- calculate_summary_distances_between_celltypes(distances)

summary_distances

An example of the interpretation of this result is: "average pairwise distance between cells of r summary_distances[4,"Reference"] and r summary_distances[4,"Target"] is r round(summary_distances[4,"Mean"],4)".

These pairwise cell distances can then be visualised as a heatmap with plot_distance_heatmap(). This example shows the average pairwise distances between cell types. Note that the pairwise distances are symmetrical (the average distance between cell type A and cell type B is the same as the average distance between cell Type B and cell Type A).

plot_distance_heatmap(phenotype_distances_result = summary_distances, metric = "mean")

This plot shows that Tumour cells are interacting most closely with Tumour cells and Immune3 cells.

Minimum cell distances

We can also calculate the minimum distances between cell types with calculate_minimum_distances_between_celltypes(). Unlike the pairwise distance where we calculate the distances between all cell types of interest, here we only identify the distance to the closest cell of type B to each of the reference cells of type A.

min_dist <- calculate_minimum_distances_between_celltypes(
  spe_object = formatted_image, 
  cell_types_of_interest = c("Tumour", "Immune1", "Immune2","Immune3", "Others"),
  feature_colname = "Cell.Type")

The minimum distances can be visualised as a violin plot with plot_cell_distances_violin(). Visualisation of this distribution often reveals whether pairs of cells are evenly spaced across the image, or whether there are clusters of pairs of cell types.

plot_cell_distances_violin(cell_to_cell_dist = min_dist)

We can also calculate summary statistics for the distances between each combination of cell types, the mean, median, min, max and standard deviation, with calculate_summary_distances_between_celltypes().

min_summary_dist <- calculate_summary_distances_between_celltypes(min_dist)

# show the first five rows
min_summary_dist[seq_len(5),]

Unlike the pairwise distance, the minimum distances are not symmetrical, and therefore we output a summary of the minimum distances specifying the reference and target cell types used.

An example of the interpretation of this result is: "average minimum distance between cells of r min_summary_dist[4,"Reference"] and r min_summary_dist[4,"Target"] is r round(min_summary_dist[4,"Mean"], 5)".

Similarly, the summary statistics of the minimum distances can also be visualised by a heatmap. This example shows the average minimum distance between cell types.

plot_distance_heatmap(phenotype_distances_result = min_summary_dist, metric = "mean")

You can access the vignettes for other modules of SPIAT here:

• Overview of SPIAT
• Data reading and formatting
• Quality control and visualisation
• Cell colocalisation
• Spatial heterogeneity
• Tissue structure
• Cellular neighborhood

Reproducibility

sessionInfo()

Author Contributions

AT, YF, TY, ML, JZ, VO, MD are authors of the package code. MD and YF wrote the vignette. AT, YF and TY designed the package.

TrigosTeam/SPIAT documentation built on Aug. 22, 2024, 7:50 p.m.