#' @name getMoHeatmap
#' @title Get multi-omics comprehensive heatmap
#' @description This function vertically concatenates multiple heatmap derived from each omics data. `getMoHeatmap` supports customized column annotation and is able to mark the selected features if indicated.
#' @param data A list of data frame or matrix storing multiple omics data with rows for features and columns for samples.
#' @param is.binary A logicial vector to indicate if the subdata is binary matrix of 0 and 1 such as mutation.
#' @param row.title A string vector to assign titles for each subdata.
#' @param legend.name A string vector to assign legend title for each subdata.
#' @param clust.res A clust.res object returned by `getMOIC()` with one specified algorithm or `get\%algorithm_name\%` or `getConsensusMOIC()` with a list of multiple algorithms.
#' @param clust.dend A dendrogram object returned returned by `getMOIC()` with one specified algorithm or `get\%algorithm_name\%` or `getConsensusMOIC()` with a list of multiple algorithms.
#' @param show.col.dend A logical vector to indicate if showing the dendrogram for column at the top of heatmap.
#' @param show.colnames A logical vector to indicate if showing the names for column at the bottom of heatmap.
#' @param show.row.dend A logical vector to indicate if showing the dendrogram for row of each subdata.
#' @param show.rownames A logical vector to indicate if showing the names for row of each subdata.
#' @param clust.dist.row A string vector to assign distance method for clustering each subdata at feature dimension.
#' @param clust.method.row A string vector to assign clustering method for clustering each subdata at feature dimension.
#' @param clust.col A string vector storing colors for annotating each subtype at the top of heatmap.
#' @param color A list of string vectors storing colors for each subheatmap of subdata.
#' @param annCol A data.frame storing annotation information for samples with exact the same sample order with data parameter.
#' @param annColors A list of string vectors for colors matched with annCol.
#' @param annRow A list of string vectors to indicate which features belong to which subdata should be annotated specifically in subheatmap.
#' @param width An integer value to indicate the width for each subheatmap with unit of cm.
#' @param height An integer value to indicate the height for each subheatmap with unit of cm.
#' @param fig.path A string value to indicate the output path for storing the comprehensive heatmap.
#' @param fig.name A string value to indicate the name of the comprehensive heatmap.
#' @return A pdf of multi-omics comprehensive heatmap
#' @importFrom ComplexHeatmap HeatmapAnnotation Heatmap rowAnnotation anno_mark draw ht_opt %v%
#' @importFrom ClassDiscovery distanceMatrix
#' @importFrom grDevices pdf dev.off colorRampPalette
#' @importFrom circlize colorRamp2
#' @importFrom dplyr %>%
#' @references Gu Z, Eils R, Schlesner M (2016). Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics.
#' @export
#' @examples # There is no example and please refer to vignette.
getMoHeatmap <- function(data = NULL,
is.binary = c(FALSE, FALSE, FALSE, FALSE, FALSE, FALSE),
row.title = c("Data1","Data2","Data3","Data4","Data5","Data6"),
legend.name = c("Data1","Data2","Data3","Data4","Data5","Data6"),
clust.res = NULL,
clust.dend = NULL,
show.col.dend = TRUE,
show.colnames = FALSE,
show.row.dend = c(TRUE, TRUE, TRUE, TRUE, TRUE, TRUE),
show.rownames = c(FALSE, FALSE, FALSE, FALSE, FALSE, FALSE),
clust.dist.row = c("pearson","pearson","pearson","pearson","pearson","pearson"),
clust.method.row = c("ward.D","ward.D","ward.D","ward.D","ward.D","ward.D"),
clust.col = c("#2EC4B6","#E71D36","#FF9F1C","#BDD5EA","#FFA5AB","#011627","#023E8A","#9D4EDD"),
color = rep(list(c("#00FF00", "#000000", "#FF0000")),length(data)),
annCol = NULL,
annColors = NULL,
annRow = NULL,
width = 6,
height = 4,
fig.path = getwd(),
fig.name = "moheatmap") {
ht_opt$message = FALSE
defaultW <- getOption("warn")
options(warn = -1)
# check data
if(is.null(names(data))){
names(data) <- sprintf("dat%s", 1:length(data))
}
n_dat <- length(data)
if(n_dat > 6){
stop('current verision of MOVICS can support up to 6 datasets.')
}
if(n_dat < 2){
stop('current verision of MOVICS needs at least 2 omics data.')
}
colvec <- clust.col[1:length(unique(clust.res$clust))]
names(colvec) <- paste0("CS",unique(clust.res$clust))
if(!is.null(annCol) & !is.null(annColors)) {
annCol <- annCol[colnames(data[[1]]), , drop = FALSE]
annCol$Subtype <- paste0("CS",clust.res[colnames(data[[1]]),"clust"])
annColors[["Subtype"]] <- colvec
if(is.null(clust.dend)) {
clust.res <- clust.res[order(clust.res$clust),]
annCol <- annCol[clust.res$samID, , drop = FALSE]
}
ha <- ComplexHeatmap::HeatmapAnnotation(df = annCol,
col = annColors,
border = FALSE)
} else {
annCol <- data.frame("Subtype" = paste0("CS",clust.res[colnames(data[[1]]),"clust"]),
row.names = colnames(data[[1]]),
stringsAsFactors = FALSE)
annColors <- list("Subtype" = colvec)
if(is.null(clust.dend)) {
clust.res <- clust.res[order(clust.res$clust),]
annCol <- annCol[clust.res$samID,,drop = FALSE]
}
ha <- ComplexHeatmap::HeatmapAnnotation(df = annCol,
col = annColors,
border = FALSE)
}
if(!is.null(annRow)) {
if(!is.list(annRow)) {stop("argument of annRow should be a list!")}
}
ht <- list()
for (i in 1:n_dat) {
hcg <- hclust(ClassDiscovery::distanceMatrix(as.matrix(t(data[[i]])), clust.dist.row[i]), clust.method.row[i])
if(is.null(annRow[[i]][1])) {
rowlab <- ""
rowlab.index <- 0
} else if (is.na(annRow[[i]][1])) {
rowlab <- ""
rowlab.index <- 0
} else {
rowlab <- intersect(rownames(data[[i]]),annRow[[i]])
rowlab.index <- match(rowlab, rownames(data[[i]]))
}
if(is.null(clust.dend)) {
data <- lapply(data, function(x) x[,clust.res$samID])
if(!is.binary[i]) {
ht[[i]] <- ComplexHeatmap::Heatmap(matrix = as.matrix(data[[i]]),
row_title = row.title[i],
name = legend.name[i],
cluster_columns = FALSE,
cluster_rows = hcg,
show_column_dend = FALSE,
show_column_names = show.colnames,
show_row_dend = show.row.dend[i],
show_row_names = show.rownames[i],
col = grDevices::colorRampPalette(color[[i]])(64),
top_annotation = switch((i == 1) + 1, NULL, ha),
width = grid::unit(width, "cm"),
height = grid::unit(height, "cm"),
heatmap_legend_param = list(at = pretty(range(data[[i]])),
labels = pretty(range(data[[i]]))),
right_annotation = ComplexHeatmap::rowAnnotation(link =
anno_mark(at = rowlab.index,
labels = rowlab,
which = "row",
lines_gp = grid::gpar(fontsize = 5),
link_width = grid::unit(3, "mm"),
padding = grid::unit(0.8, "mm"),
labels_gp = grid::gpar(fontsize = 7))))
} else {
col_fun = circlize::colorRamp2(c(0, 1), color[[i]])
ht[[i]] <- ComplexHeatmap::Heatmap(matrix = as.matrix(data[[i]]),
row_title = row.title[i],
name = legend.name[i],
cluster_columns = FALSE,
cluster_rows = hcg,
show_column_dend = FALSE,
show_column_names = show.colnames,
show_row_dend = show.row.dend[i],
show_row_names = show.rownames[i],
col = color[[i]],
top_annotation = switch((i == 1) + 1, NULL, ha),
width = grid::unit(width, "cm"),
height = grid::unit(height, "cm"),
heatmap_legend_param = list(at = c(0, 1),
legend_gp = grid::gpar(fill = col_fun(c(0,1))),
labels = c("0", "1")),
right_annotation = ComplexHeatmap::rowAnnotation(link =
anno_mark(at = rowlab.index,
labels = rowlab,
which = "row",
lines_gp = grid::gpar(fontsize = 5),
link_width = grid::unit(3, "mm"),
padding = grid::unit(0.8, "mm"),
labels_gp = grid::gpar(fontsize = 7))))
}
} else {
if(!is.binary[i]) {
ht[[i]] <- ComplexHeatmap::Heatmap(matrix = as.matrix(data[[i]]),
row_title = row.title[i],
name = legend.name[i],
cluster_columns = clust.dend,
cluster_rows = hcg,
show_column_dend = show.col.dend,
show_column_names = show.colnames,
show_row_dend = show.row.dend[i],
show_row_names = show.rownames[i],
col = grDevices::colorRampPalette(color[[i]])(64),
top_annotation = switch((i == 1) + 1, NULL, ha),
width = grid::unit(width, "cm"),
height = grid::unit(height, "cm"),
heatmap_legend_param = list(at = pretty(range(data[[i]])),
labels = pretty(range(data[[i]]))),
right_annotation = ComplexHeatmap::rowAnnotation(link =
anno_mark(at = rowlab.index,
labels = rowlab,
which = "row",
lines_gp = grid::gpar(fontsize = 5),
link_width = grid::unit(3, "mm"),
padding = grid::unit(0.8, "mm"),
labels_gp = grid::gpar(fontsize = 7))))
} else {
col_fun = circlize::colorRamp2(c(0, 1), color[[i]])
ht[[i]] <- ComplexHeatmap::Heatmap(matrix = as.matrix(data[[i]]),
row_title = row.title[i],
name = legend.name[i],
cluster_columns = clust.dend,
cluster_rows = hcg,
show_column_dend = show.col.dend,
show_column_names = show.colnames,
show_row_dend = show.row.dend[i],
show_row_names = show.rownames[i],
col = color[[i]],
top_annotation = switch((i == 1) + 1, NULL, ha),
width = grid::unit(width, "cm"),
height = grid::unit(height, "cm"),
heatmap_legend_param = list(at = c(0, 1),
legend_gp = grid::gpar(fill = col_fun(c(0,1))),
labels = c("0", "1")),
right_annotation = ComplexHeatmap::rowAnnotation(link =
anno_mark(at = rowlab.index,
labels = rowlab,
which = "row",
lines_gp = grid::gpar(fontsize = 5),
link_width = grid::unit(3, "mm"),
padding = grid::unit(0.8, "mm"),
labels_gp = grid::gpar(fontsize = 7))))
}
}
}
if(n_dat == 1){
ht_list <- ht[[1]]
}
if(n_dat == 2){
ht_list <- ht[[1]] %v% ht[[2]]
}
if(n_dat == 3){
ht_list <- ht[[1]] %v% ht[[2]] %v% ht[[3]]
}
if(n_dat == 4){
ht_list <- ht[[1]] %v% ht[[2]] %v% ht[[3]] %v% ht[[4]]
}
if(n_dat == 5){
ht_list <- ht[[1]] %v% ht[[2]] %v% ht[[3]] %v% ht[[4]] %v% ht[[5]]
}
if(n_dat == 6){
ht_list <- ht[[1]] %v% ht[[2]] %v% ht[[3]] %v% ht[[4]] %v% ht[[5]] %v% ht[[6]]
}
outFile <- file.path(fig.path,paste0(fig.name,".pdf"))
if(is.null(annCol)) {
pdf(outFile, width = width, height = height * n_dat/2)
} else {
pdf(outFile, width = width, height = height * n_dat/1.5)
}
draw(ht_list, merge_legend = TRUE, heatmap_legend_side = "right") # output to pdf
invisible(dev.off())
draw(ht_list, merge_legend = TRUE, heatmap_legend_side = "right") # output to screen
options(warn = defaultW)
}
Datasets
Models
