In drieslab/Giotto_site_suite: Spatial Single-Cell Transcriptomics Toolbox

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(Giotto)

Install Python modules

To run this vignette you need to install all the necessary Python modules.

1. This can be done manually, see https://rubd.github.io/Giotto_site/articles/installation_issues.html#python-manual-installation

2. This can be done within R using our installation tools (installGiottoEnvironment), see https://rubd.github.io/Giotto_site/articles/tut0_giotto_environment.html for more information.

(optional) set giotto instructions

# to automatically save figures in save_dir set save_plot to TRUE
temp_dir = '~/Temp/'
myinstructions = createGiottoInstructions(save_dir = temp_dir,
                                          save_plot = FALSE, 
                                          show_plot = F)

1. Create a Giotto object

minimum requirements:
- matrix with expression information (or path to)
- x,y(,z) coordinates for cells or spots (or path to)

# giotto object 
expr_path = system.file("extdata", "starmap_expr.txt.gz", package = 'Giotto')
loc_path = system.file("extdata", "starmap_cell_loc.txt", package = 'Giotto')
starmap_mini <- createGiottoObject(raw_exprs = expr_path,
                                   spatial_locs = loc_path,
                                   instructions = myinstructions)

How to work with Giotto instructions that are part of your Giotto object:
- show the instructions associated with your Giotto object with showGiottoInstructions
- change one or more instructions with changeGiottoInstructions
- replace all instructions at once with replaceGiottoInstructions
- read or get a specific giotto instruction with readGiottoInstructions
Of note, the python path can only be set once in an R session. See the reticulate package for more information.

# show instructions associated with giotto object (starmap_mini)
showGiottoInstructions(starmap_mini)

2. processing steps

• filter genes and cells based on detection frequencies
• normalize expression matrix (log transformation, scaling factor and/or z-scores)
• add cell and gene statistics (optional)
• adjust expression matrix for technical covariates or batches (optional). These results will be stored in the custom slot.

filterDistributions(starmap_mini, detection = 'genes')
filterDistributions(starmap_mini, detection = 'cells')
filterCombinations(starmap_mini,
                   expression_thresholds = c(1),
                   gene_det_in_min_cells = c(50, 100, 200),
                   min_det_genes_per_cell = c(20, 28, 28))

starmap_mini <- filterGiotto(gobject = starmap_mini,
                           expression_threshold = 1,
                           gene_det_in_min_cells = 50,
                           min_det_genes_per_cell = 20,
                           expression_values = c('raw'),
                           verbose = T)
starmap_mini <- normalizeGiotto(gobject = starmap_mini, scalefactor = 6000, verbose = T)
starmap_mini <- addStatistics(gobject = starmap_mini)

3. dimension reduction

• identify highly variable genes (HVG) will not be performed here, because there are only few genes
• perform PCA
• identify number of significant prinicipal components (PCs)
• run UMAP and/or TSNE on PCs (or directly on matrix)

starmap_mini <- runPCA(gobject = starmap_mini, method = 'factominer')
screePlot(starmap_mini, ncp = 30)
plotPCA(gobject = starmap_mini)

# 2D umap
starmap_mini <- runUMAP(starmap_mini, dimensions_to_use = 1:8)
plotUMAP(gobject = starmap_mini)

# 3D umap
starmap_mini <- runUMAP(starmap_mini, dimensions_to_use = 1:8, name = '3D_umap', n_components = 3)
plotUMAP_3D(gobject = starmap_mini, dim_reduction_name = '3D_umap')

# 2D tsne
starmap_mini <- runtSNE(starmap_mini, dimensions_to_use = 1:8)
plotTSNE(gobject = starmap_mini)

4. clustering

• create a shared (default) nearest network in PCA space (or directly on matrix)
• cluster on nearest network with Leiden or Louvan (kmeans and hclust are alternatives)

starmap_mini <- createNearestNetwork(gobject = starmap_mini, dimensions_to_use = 1:8, k = 25)
starmap_mini <- doLeidenCluster(gobject = starmap_mini, resolution = 0.5, n_iterations = 1000)

# 2D umap
plotUMAP(gobject = starmap_mini, cell_color = 'leiden_clus', show_NN_network = T, point_size = 2.5)

# 3D umap
plotUMAP_3D(gobject = starmap_mini, dim_reduction_name = '3D_umap', cell_color = 'leiden_clus')

# 2D umap + coordinates
spatDimPlot(gobject = starmap_mini, cell_color = 'leiden_clus',
            dim_point_size = 2, spat_point_size = 2.5)

# 3D umap + coordinates
spatDimPlot3D(gobject = starmap_mini, cell_color = 'leiden_clus', dim_reduction_name = '3D_umap')

# heatmap and dendrogram
showClusterHeatmap(gobject = starmap_mini, cluster_column = 'leiden_clus')
showClusterDendrogram(starmap_mini, h = 0.5, rotate = T, cluster_column = 'leiden_clus')

5. differential expression

gini_markers = findMarkers_one_vs_all(gobject = starmap_mini,
                                                  method = 'gini',
                                                  expression_values = 'normalized',
                                                  cluster_column = 'leiden_clus',
                                                  min_genes = 20,
                                                  min_expr_gini_score = 0.5,
                                                  min_det_gini_score = 0.5)

# get top 2 genes per cluster and visualize with violinplot
topgenes_gini = gini_markers[, head(.SD, 2), by = 'cluster']
violinPlot(starmap_mini, genes = topgenes_gini$feats, cluster_column = 'leiden_clus')

# get top 6 genes per cluster and visualize with heatmap
topgenes_gini2 = gini_markers[, head(.SD, 6), by = 'cluster']
plotMetaDataHeatmap(starmap_mini, selected_genes = topgenes_gini2$feats ,
                    metadata_cols = c('leiden_clus'))

6. A. cell type annotation

clusters_cell_types = c('cell A', 'cell B', 'cell C', 'cell D',
                        'cell E', 'cell F', 'cell G', 'cell H')
names(clusters_cell_types) = 1:8
starmap_mini = annotateGiotto(gobject = starmap_mini, 
                              annotation_vector = clusters_cell_types, 
                              cluster_column = 'leiden_clus', 
                              name = 'cell_types')
# check new cell metadata
pDataDT(starmap_mini)

# visualize annotations
spatDimPlot(gobject = starmap_mini, cell_color = 'cell_types', 
            spat_point_size = 2, dim_point_size = 2)

6. B. cell type gene expression

dimGenePlot3D(starmap_mini,
              dim_reduction_name = '3D_umap',
              expression_values = 'scaled',
              genes = "Pcp4",
              genes_high_color = 'red', genes_mid_color = 'white', genes_low_color = 'darkblue')

spatGenePlot3D(starmap_mini,
               expression_values = 'scaled',
               genes = "Pcp4",
               show_other_cells = F,
               genes_high_color = 'red', genes_mid_color = 'white', genes_low_color = 'darkblue')

7. spatial grid

Create a grid based on defined stepsizes in the x,y(,z) axes.

starmap_mini <- createSpatialGrid(gobject = starmap_mini,
                              sdimx_stepsize = 200,
                              sdimy_stepsize = 200,
                              sdimz_stepsize = 20,
                              minimum_padding = 10)
showGrids(starmap_mini)

# visualize grid
spatPlot2D(gobject = starmap_mini, show_grid = T, point_size = 1.5)

8. spatial network

Only the method = delaunayn_geometry can make 3D Delaunay networks. This requires the package geometry to be installed.

• visualize information about the default Delaunay network
• create a spatial Delaunay network (default)
• create a spatial kNN network

plotStatDelaunayNetwork(gobject = starmap_mini, maximum_distance = 200, 
                        method = 'delaunayn_geometry')
starmap_mini = createSpatialNetwork(gobject = starmap_mini,
                                    minimum_k = 2, 
                                    maximum_distance_delaunay = 200, 
                                    method = 'Delaunay', 
                                    delaunay_method = 'delaunayn_geometry')
starmap_mini = createSpatialNetwork(gobject = starmap_mini,
                                    minimum_k = 2, 
                                    method = 'kNN', k = 10)
showNetworks(starmap_mini)

# visualize the two different spatial networks  
spatPlot(gobject = starmap_mini, show_network = T,
         network_color = 'blue', spatial_network_name = 'Delaunay_network',
         point_size = 2.5, cell_color = 'leiden_clus')

spatPlot(gobject = starmap_mini, show_network = T,
         network_color = 'blue', spatial_network_name = 'kNN_network',
         point_size = 2.5, cell_color = 'leiden_clus')

9. spatial genes

Identify spatial genes with 3 different methods:
- binSpect with kmeans binarization (default)
- binSpect with rank binarization
- silhouetteRank

Visualize top 4 genes per method.

km_spatialgenes = binSpect(starmap_mini)
spatGenePlot(starmap_mini, expression_values = 'scaled', 
             genes = km_spatialgenes[1:4]$feats,
             point_shape = 'border', point_border_stroke = 0.1,
             show_network = F, network_color = 'lightgrey', point_size = 2.5,
             cow_n_col = 2)

rank_spatialgenes = binSpect(starmap_mini, bin_method = 'rank')
spatGenePlot(starmap_mini, expression_values = 'scaled', 
             genes = rank_spatialgenes[1:4]$feats,
             point_shape = 'border', point_border_stroke = 0.1,
             show_network = F, network_color = 'lightgrey', point_size = 2.5,
             cow_n_col = 2)

silh_spatialgenes = silhouetteRank(gobject = starmap_mini) # TODO: suppress print output
spatGenePlot(starmap_mini, expression_values = 'scaled', 
             genes = silh_spatialgenes[1:4]$genes,
             point_shape = 'border', point_border_stroke = 0.1,
             show_network = F, network_color = 'lightgrey', point_size = 2.5,
             cow_n_col = 2)

10. spatial co-expression patterns

Identify robust spatial co-expression patterns using the spatial network or grid and a subset of individual spatial genes.
1. calculate spatial correlation scores
2. cluster correlation scores

# 1. calculate spatial correlation scores 
ext_spatial_genes = km_spatialgenes[1:20]$feats
spat_cor_netw_DT = detectSpatialCorGenes(starmap_mini,
                                         method = 'network', 
                                         spatial_network_name = 'Delaunay_network',
                                         subset_genes = ext_spatial_genes)

# 2. cluster correlation scores
spat_cor_netw_DT = clusterSpatialCorGenes(spat_cor_netw_DT, 
                                          name = 'spat_netw_clus',
                                          k = 6)
heatmSpatialCorGenes(starmap_mini,
                     spatCorObject = spat_cor_netw_DT, 
                     use_clus_name = 'spat_netw_clus')

netw_ranks = rankSpatialCorGroups(starmap_mini, 
                                  spatCorObject = spat_cor_netw_DT, 
                                  use_clus_name = 'spat_netw_clus')
top_netw_spat_cluster = showSpatialCorGenes(spat_cor_netw_DT, 
                                            use_clus_name = 'spat_netw_clus',
                                            selected_clusters = 6, 
                                            show_top_genes = 1)

cluster_genes_DT = showSpatialCorGenes(spat_cor_netw_DT, 
                                       use_clus_name = 'spat_netw_clus',
                                       show_top_genes = 1)
cluster_genes = cluster_genes_DT$clus; names(cluster_genes) = cluster_genes_DT$feat_ID


starmap_mini = createMetagenes(starmap_mini, gene_clusters = cluster_genes, name = 'cluster_metagene')
spatCellPlot(starmap_mini,
             spat_enr_names = 'cluster_metagene',
             cell_annotation_values = netw_ranks$clusters,
             point_size = 1.5, cow_n_col = 3)

11. spatial HMRF domains

hmrf_folder = paste0(temp_dir,'/','11_HMRF/')
if(!file.exists(hmrf_folder)) dir.create(hmrf_folder, recursive = T)

# perform hmrf
my_spatial_genes = km_spatialgenes[1:20]$feats
HMRF_spatial_genes = doHMRF(gobject = starmap_mini,
                            expression_values = 'scaled',
                            spatial_genes = my_spatial_genes,
                            spatial_network_name = 'Delaunay_network',
                            k = 6,
                            betas = c(10,2,2),
                            output_folder = paste0(hmrf_folder, '/', 'Spatial_genes/SG_top20_k6_scaled'))

# check and select hmrf
for(i in seq(10, 14, by = 2)) {
  viewHMRFresults2D(gobject = starmap_mini,
                    HMRFoutput = HMRF_spatial_genes,
                    k = 6, betas_to_view = i,
                    point_size = 2)
}

starmap_mini = addHMRF(gobject = starmap_mini,
                  HMRFoutput = HMRF_spatial_genes,
                  k = 6, betas_to_add = c(12),
                  hmrf_name = 'HMRF')

# visualize selected hmrf result
giotto_colors = Giotto:::getDistinctColors(6)
names(giotto_colors) = 1:6
spatPlot(gobject = starmap_mini, cell_color = 'HMRF_k6_b.12',
         point_size = 3, coord_fix_ratio = 1, cell_color_code = giotto_colors)

12. cell neighborhood: cell-type/cell-type interactions

set.seed(seed = 2841)
cell_proximities = cellProximityEnrichment(gobject = starmap_mini,
                                           cluster_column = 'cell_types',
                                           spatial_network_name = 'Delaunay_network',
                                           adjust_method = 'fdr',
                                           number_of_simulations = 1000)
# barplot
cellProximityBarplot(gobject = starmap_mini, 
                     CPscore = cell_proximities, 
                     min_orig_ints = 5, min_sim_ints = 5)
## heatmap
cellProximityHeatmap(gobject = starmap_mini, CPscore = cell_proximities, 
                     order_cell_types = T, scale = T,
                     color_breaks = c(-1.5, 0, 1.5), 
                     color_names = c('blue', 'white', 'red'))

# network
cellProximityNetwork(gobject = starmap_mini, CPscore = cell_proximities, 
                     remove_self_edges = T, only_show_enrichment_edges = T)

# network with self-edges
cellProximityNetwork(gobject = starmap_mini, CPscore = cell_proximities,
                     remove_self_edges = F, self_loop_strength = 0.3,
                     only_show_enrichment_edges = F,
                     rescale_edge_weights = T,
                     node_size = 8,
                     edge_weight_range_depletion = c(1, 2),
                     edge_weight_range_enrichment = c(2,5))

visualization of specific cell types

pDataDT(starmap_mini)
# Option 1
spec_interaction = "cell D--cell H" # needs to be in alphabetic order! first D, then H
cellProximitySpatPlot2D(gobject = starmap_mini,
                        interaction_name = spec_interaction,
                        show_network = T,
                        cluster_column = 'cell_types',
                        cell_color = 'cell_types',
                        cell_color_code = c('cell H' = 'lightblue', 'cell D' = 'red'),
                        point_size_select = 4, point_size_other = 2)

# Option 2: create additional metadata
starmap_mini = addCellIntMetadata(starmap_mini,
                             spatial_network = 'Delaunay_network',
                             cluster_column = 'cell_types',
                             cell_interaction = spec_interaction,
                             name = 'D_H_interactions')
spatPlot(starmap_mini, cell_color = 'D_H_interactions', legend_symbol_size = 3,
         select_cell_groups =  c('other_cell D', 'other_cell H', 'select_cell D', 'select_cell H'))

13. 2D cross sections from 3D object

# create cross section
starmap_mini = createCrossSection(starmap_mini,
                                method="equation",
                                equation=c(0,1,0,600),
                                extend_ratio = 0.6)

# show cross section
insertCrossSectionSpatPlot3D(starmap_mini, cell_color = 'leiden_clus',
                             axis_scale = 'cube',
                             point_size = 2)

insertCrossSectionGenePlot3D(starmap_mini, expression_values = 'scaled',
                             axis_scale = "cube",
                             genes = "Slc17a7")


# for cell annotation
crossSectionPlot(starmap_mini,
                 point_size = 2, point_shape = "border",
                 cell_color = "leiden_clus")

crossSectionPlot3D(starmap_mini,
                   point_size = 2, cell_color = "leiden_clus",
                   axis_scale = "cube")


# for gene expression
crossSectionGenePlot(starmap_mini,
                     genes = "Slc17a7",
                     point_size = 2,
                     point_shape = "border",
                     cow_n_col = 1.5,
                     expression_values = 'scaled')

crossSectionGenePlot3D(starmap_mini,
                       point_size = 2,
                       genes = c("Slc17a7"),
                       expression_values = 'scaled')

14. export Giotto Analyzer to Viewer

viewer_folder = paste0(temp_dir, '/', 'Mouse_cortex_viewer')

# select annotations, reductions and expression values to view in Giotto Viewer
exportGiottoViewer(gobject = starmap_mini, output_directory = viewer_folder,
                   factor_annotations = c('cell_types',
                                          'leiden_clus',
                                          'HMRF_k6_b.12'),
                   numeric_annotations = 'total_expr',
                   dim_reductions = c('umap'),
                   dim_reduction_names = c('umap'),
                   expression_values = 'scaled',
                   expression_rounding = 3,
                   overwrite_dir = T)

drieslab/Giotto_site_suite documentation built on April 26, 2023, 11:51 p.m.