IGAN

A method for Inferring cell-cell communication pathways represented by spatial gene associations based on spatial transcriptomic data

Installation

IGAN can be installed in R by devtools:

devtools::install_github('Zhu-JC/IGAN')

Tutorials

library(IGAN2)
library(clusterProfiler)
library(org.Hs.eg.db)
library(org.Mm.eg.db)
library(ggplot2)
library(ggsankey)
library(networkD3)
library(KEGGREST)
library(tidyverse)
library(plotly)
library(magick)
library(CellChat)
library(patchwork)
library(NMF)
library(ggalluvial)
library(ComplexHeatmap)
library(gridExtra)

#loading input datas.
##spatial is coordinate information, a dataframe which rows represent spots and 2 cols represent coordinate.
spatial = read.csv('spatial.csv')

##ident is celltype information, a 1 column dataframe which rows represent spots.
ident = read.csv('ident.csv')

##M is gene expression matrix, a dataframe which cols represent genes and rows represent spots.
M = read.csv('M')

##gene is genes' name of the dataset, a 1 column dataframe which rows represent genes.
gene = read.csv('gene.csv')

##genelist1 is gene list selected in sending spots, a 1 column dataframe which rows represent genes.
genelist1 = read.csv('genelist1')

##genelist2 is gene list selected in receiving spots, a 1 column dataframe which rows represent genes.
genelist2 = read.csv('genelist2')

gene2 = toupper(gene[,1])
gene_list1 = toupper(gene_list1[,1])
gene_list2 = toupper(gene_list2[,1])
match_list1 = na.omit(match(gene_list1, gene2))
match_list2 = na.omit(match(gene_list2, gene2))

#Grouping spots pairs into groups.
celltype = as.data.frame(c('CAF','HSC','Kupffer','LSEC','T','hepatocyte 1','hepatocyte 2','hepatocyte 3','interferon response','monocyte_DC','vascular smooth mc','tumor I','tumor II')) cells_group = cells_to_group(spatial, ident, 5000, celltype)

#Compute gene-gene associations in every single spots pair.
result = IGAN(match_list1, match_list2, gene, cells_group, M, 0.01)
result is a list of n lists, n is the number of groups get in cells_to_group. Each list in result containing m dataframes, m is the number of genes corresponding gene_list1. In each dataframe, rows represent target genes' order which corresponding gene_list2 and cols represent the order of cell pairs within the corresponding groups. 0 represents no association and 1 represents having association.

#Loading the precomputed tumor data.
load('tumor.Rdata')

#Get the matrix contain every spots' CCC feature.
cccM = get_cccM(result, cells_group, spatial, gene, match_list1, match_list2)
cccM is described in network analysis of IGAN part in our paper.

Gene_corM = getGene_corM(result, cells_group, gene, match_list1, match_list2)

#Get the graph of CCC activity of every spot.
a = plot_cor(result, cells_group, spatial)
a + scale_color_distiller(palette = "Set1")

#Get the graph of SPP1's CCC activity of every spot.
a = plot_Genecor('SPP1', result, cells_group, spatial, match_list1, gene)
a + scale_color_distiller(palette = "Set1")

plot_ligand_target(result, spatial, cells_group, gene, match_list1, match_list2, 'Plg')

#Get the cellchat's network circle graph.
par(mfrow = c(1,1), xpd=TRUE)
plot_cellchat('SPP1',result,cells_group,gene,match_list1,ident)
dev.off()

#Get the sankey graph of ligand's GO-ligand-recepter-downstream pathway.
OrgDb = org.Mm.eg.db sankey_graph = plot_sankey(result,cells_group,gene,match_list1,match_list2,0.05,5,OrgDb = OrgDb,send_cell = 'monocyte_DC',rec_cell = 'CAF',threshold = 10)
sankeyNetwork(Links = sankey_graph[[1]], Nodes =sankey_graph[[2]], Source = 'source', Target = 'target', Value = 'value', NodeID = 'name',LinkGroup = 'color', units = 'TWh', fontSize = 20, nodeWidth = 30, margin = 1, height = NULL, sinksRight = FALSE)
dev.off()

#Get the graph of CCC pattern.
pattern_graph = plot_pattern(result,cells_group,gene,match_list1,match_list2,ident,'outgoing',4)
pushViewport(viewport(x = 0.1, y = 0.1, width = 0.2, height = 0.8, just = c("left", "bottom")))
grid.draw(pattern_graph[[1]])
popViewport()
pushViewport(viewport(x = 0.6, y = 0.1, width = 0.2, height = 0.8, just = c("left", "bottom")))
grid.draw(pattern_graph[[2]])
popViewport()