scCrossTalk

Infer cell-cell communications based on CellTalkDB

Install

# download the source package of scCrossTalk-1.0.tar.gz and install it
# ensure the right directory for scCrossTalk-1.0.tar.gz
install.packages(pkgs = 'scCrossTalk-1.0.tar.gz',repos = NULL, type = "source")

or

# install devtools and install scCrossTalk
install.packages(pkgs = 'devtools')
devtools::install_github('ZJUFanLab/scCrossTalk')

Usage

library(scCrossTalk)

Find highly expressed ligand-receptor pairs between pairwise clusters

Find highly expressed ligands and receptors between pairwise clusters using Z score for a Seurat object (>= 3.0.0) after log1p normalization, cluster analysis and tSNE or Umap dimensionality reduction

clu_pairs <- FindPairs(object = mouse_kidney_203_Seurat,
                       species = "Mouse",
                       use_LRdb = "LRdb",
                       revise_gene = T,
                       cell_min_pct = 0.25,
                       p_value = 0.05)

Find significant LR pairs between pairwise clusters

PairsSig(clu_pairs = clu_pairs, per_num = 1000, pvalue = 0.05)

Find significant crosstalk between pairwise clusters

CrossTalkSig(clu_pairs = clu_pairs, per_num = 1000, pvalue = 0.05)

Plot LR pairs between pairwise clusters

PlotPairsNode(clu_pairs = clu_pairs,
              show_sig = F,
              edge_width = 1,
              edge_alpha = 0.5,
              node_size_min = 1,
              node_size_max = 10,
              text_size = 3)

PlotPairsNet(clu_pairs = clu_pairs,
             show_sig = F,
             show_clu_node = T,
             layout = "nicely",
             show_text_cutoff = 0,
             node_size_min = 5,
             node_size_max = 10,
             text_size = 3,
             text_col = "black",
             edge_width = 0.5,
             edge_col = "black",
             edge_alpha = 0.2)

PlotPairsHeat(clu_pairs = clu_pairs,
              LR_pairs = "all",
              show_clusters = "all",
              show_sig = F,
              color_low = "white",
              color_high = "red",
              border_color = "grey60",
              cluster_rows = T,
              cluster_cols = T,
              symbol = "*",
              symbol_col = "black",
              symbol_size = 12)

PlotPairsBubble(clu_pairs = clu_pairs,
                LR_pairs = "all",
                show_clusters = "all",
                if_directed = T, 
                show_sig = F,
                bubble_col = "black",
                bubble_alpha = 0.6,
                bubble_max_size = 20,
                show_text_cutoff = 1)

PlotPairsCircle(clu_pairs = clu_pairs,
                show_sig = F,
                ligand_clu = "1",
                receptor_clu = "2")

PlotPairsDim(clu_pairs = clu_pairs,
             ligand = "Apoe",
             ligand_clu = "4",
             receptor = "Sdc4",
             receptor_clu = "2",
             reduction = "umap",
             size = 1,
             text_size = 12)

PlotPairsViolin(clu_pairs = clu_pairs,
                ligand = "Apoe",
                ligand_clu = "4",
                receptor = "Sdc4",
                receptor_clu = "2",
                reduction = "umap",
                show_jitter = T,
                jitter_size = 2)

Plot crosstalk between pairwise clusters

PlotCrossTalkSan(clu_pairs = clu_pairs, show_type = "number", show_sig = F)

PlotCrossTalkCircle(clu_pairs = clu_pairs, show_type = "score", show_sig = F)

PlotCrossTalkHeat(clu_pairs = clu_pairs,
                  show_type = "number",
                  if_directed = T,
                  show_sig = F,
                  color_low = "white",
                  color_high = "red",
                  border_color = "grey60",
                  cluster_rows = T,
                  cluster_cols = T,
                  symbol = "*",
                  symbol_col = "black",
                  symbol_size = 12)

PlotCrossTalkNet(clu_pairs = NULL,
                 show_type = NULL,
                 layout = "nicely",
                 show_sig = F,
                 edge_col = "black",
                 edge_alpha = 0.1,
                 node_size_min = 5,
                 node_size_max = 10,
                 text_size = 3,
                 text_col = "black")