library(methods)
suppressPackageStartupMessages(library(GetoptLong))
type = "neg"
cutoff = 0.01
meandiff = 0.3
rerun = FALSE
GetoptLong("type=s", "neg|pos",
"cutoff=f", "0.05",
"meandiff=f", "0",
"rerun!", "rerun")
BASE_DIR = "/icgc/dkfzlsdf/analysis/B080/guz/roadmap_analysis/re_analysis"
source(qq("@{BASE_DIR}/scripts/configure/roadmap_configure.R"))
km = readRDS(qq("@{OUTPUT_DIR}/rds/mat_all_cr_enriched_to_gene_extend_50000_target_0.33_row_order_km3_and_km4.rds"))[[2]]
km_col = structure(brewer.pal(9, "Set1")[c(3,4,5,1)], names = c(1:4))
neg_cr_all = readRDS(qq("@{OUTPUT_DIR}/rds/all_neg_cr_w6s3.rds"))
pos_cr_all = readRDS(qq("@{OUTPUT_DIR}/rds/all_pos_cr_w6s3.rds"))
cr_all = c(neg_cr_all, pos_cr_all)
cr_all = copy_cr_attribute(neg_cr_all, cr_all)
if(type == "neg") {
cr = readRDS(qq("@{OUTPUT_DIR}/rds/all_neg_cr_w6s3_fdr_less_than_@{cutoff}_methdiff_larger_than_@{meandiff}.rds"))
cr_vice = readRDS(qq("@{OUTPUT_DIR}/rds/all_pos_cr_w6s3_fdr_less_than_@{cutoff}_methdiff_larger_than_@{meandiff}.rds"))
} else if(type == "pos") {
cr = readRDS(qq("@{OUTPUT_DIR}/rds/all_pos_cr_w6s3_fdr_less_than_@{cutoff}_methdiff_larger_than_@{meandiff}.rds"))
cr_vice = readRDS(qq("@{OUTPUT_DIR}/rds/all_neg_cr_w6s3_fdr_less_than_@{cutoff}_methdiff_larger_than_@{meandiff}.rds"))
}
sample_id = attr(cr, "sample_id")
gene = genes(TXDB)
gene = gene[intersect(names(gene), names(km))]
qqcat("extracting gene tss\n")
tss = promoters(gene, upstream = 1, downstream = 0)
tss = tss[names(tss) %in% cr$gene_id]
expr = EXPR[names(tss), , drop = FALSE]
cgi_extend = CGI; start(cgi_extend) = start(CGI) - 5000; end(cgi_extend) = end(CGI) + 5000
# there should only be one tss in +-5kb of CGI and one tss should
# only overlaps to one extended CGI
mtch = as.matrix(findOverlaps(cgi_extend, tss))
t1 = table(mtch[, 1])
t2 = table(mtch[, 2])
s1 = as.numeric(names(t1[t1 == 1]))
s2 = as.numeric(names(t2[t2 == 1]))
l = mtch[, 1] %in% s1 & mtch[, 2] %in% s2
mtch = mtch[l, ]
cgi2 = CGI[mtch[, 1]]
cgi2$gene_id = names(tss[mtch[, 2]])
names(cgi2) = cgi2$gene_id
strand(cgi2) = strand(tss[mtch[, 2]])
extend = 5000
target_ratio = mean(width(cgi2))/(extend*2 + mean(width(cgi2)))
mat = normalizeToMatrix(cr, cgi2, extend = extend, w = 50, trim = 0, mean_mode = "absolute",
mapping_column = "gene_id", target_ratio = target_ratio)
if(type == "neg") {
mat[mat == 1] = -1
}
mat_vice = normalizeToMatrix(cr_vice, cgi2, extend = extend, w = 50, trim = 0, mean_mode = "absolute",
mapping_column = "gene_id", target_ratio = target_ratio)
if(type == "pos") {
mat_vice[mat_vice == 1] = -1
}
l = rowSums(abs(mat)) > 0
mat = mat[l, ]
cgi2 = cgi2[l]
mat_vice = mat_vice[l, ]
qqcat("@{length(cgi2)}/@{length(CGI)} CGIs remains\n")
mat_corr = normalizeToMatrix(cr_all, cgi2, mapping_column = "gene_id", value_column = "corr",
extend = extend, mean_mode = "absolute", w = 50, target_ratio = target_ratio, trim = 0, empty_value = 0)
# n_tss = countOverlaps(cgi_extend, tss)
# dist = distanceToNearest(cgi2, tss)
strd = as.vector(strand(cgi2))
strd = factor(paste0(strd, "strand"), levels = c("-strand", "+strand"))
km = km[names(cgi2)]
cgi_width = width(cgi2)
cgi_width[cgi_width > quantile(cgi_width, 0.99)] = quantile(cgi_width, 0.99)
sample_id_subgroup1 = intersect(sample_id, rownames(SAMPLE[SAMPLE$subgroup == "subgroup1", ]))
sample_id_subgroup2 = intersect(sample_id, rownames(SAMPLE[SAMPLE$subgroup == "subgroup2", ]))
rdata_file = qq("@{OUTPUT_DIR}/rds/cr_enrichedheatmap_cgi_by_gene_@{type}_fdr_@{cutoff}_methdiff_@{meandiff}.RData")
if(file.exists(rdata_file) && !rerun) {
load(rdata_file)
} else {
meth_mat = enrich_with_methylation(cgi2, sample_id, target_ratio = target_ratio, extend = extend)
meth_mat[attr(meth_mat, "failed_rows"), ] = 0.5
meth_mat_1 = enrich_with_methylation(cgi2, sample_id_subgroup1, target_ratio = target_ratio, extend = extend)
failed_rows = attr(meth_mat_1, "failed_rows")
qqcat("There are @{length(failed_rows)} failed rows when normalizing methylation to the targets.\n")
meth_mat_1[failed_rows, ] = 0.5
meth_mat_2 = enrich_with_methylation(cgi2, sample_id_subgroup2, target_ratio = target_ratio, extend = extend)
failed_rows = attr(meth_mat_2, "failed_rows")
qqcat("There are @{length(failed_rows)} failed rows when normalizing methylation to the targets.\n")
meth_mat_2[failed_rows, ] = 0.5
meth_mat_diff = meth_mat_1 - meth_mat_2
cor_mat_list = list()
hist_mat_list = list()
hist_mat_list_subgroup1 = list()
hist_mat_list_subgroup2 = list()
hist_mat_list_diff = list()
for(k in seq_along(MARKS)) {
hm_sample = intersect(sample_id, chipseq_hooks$sample_id(MARKS[k]))
# applied to each sample, each mark
lt = enrich_with_histone_mark(cgi2, sample_id = sample_id, mark = MARKS[k], return_arr = TRUE, target_ratio = target_ratio, extend = extend)
arr = lt[[1]]
# only calculate the correlation when there are enough samples
if(length(hm_sample) >= 5) {
# detect regions that histone MARKS correlate to expression
expr2 = EXPR[cgi2$gene_id, intersect(colnames(EXPR), hm_sample)]
cor_mat = matrix(nrow = nrow(expr2), ncol = ncol(mat))
cor_p_mat = cor_mat
counter = set_counter(nrow(cor_mat))
for(i in seq_len(nrow(cor_mat))) {
counter()
for(j in seq_len(ncol(cor_mat))) {
x = cor(arr[i, j, ], expr2[i, ], method = "spearman")
cor_mat[i, j] = x
# cor_p_mat[i, j] = cor.test(arr[i, j, ], expr2[i, ], method = "spearman")$p.value
}
}
cat("\n")
cor_mat[is.na(cor_mat)] = 0
# cor_fdr_mat = p.adjust(cor_p_mat, method = "BH")
# l1 = cor_fdr_mat < 0.1 & cor_mat > 0
# cor_mat[l1] = 1
# l2 = cor_fdr_mat < 0.1 & cor_mat < 0
# cor_mat[l2] = -1
# cor_mat[!(l1 | l2)] = 0
cor_mat = copyAttr(mat, cor_mat)
cor_mat_list[[k]] = cor_mat
if(sum(abs(cor_mat)) == 0) {
cor_mat_list[[k]] = NA
}
} else {
cor_mat_list[[k]] = NA
}
hist_mat_list[[k]] = lt[[2]]
hist_mat_list_subgroup1[[k]] = apply(arr[, , intersect(hm_sample, sample_id_subgroup1)], c(1, 2), mean, na.rm = TRUE)
hist_mat_list_subgroup1[[k]] = copyAttr(mat, hist_mat_list_subgroup1[[k]])
hist_mat_list_subgroup2[[k]] = apply(arr[, , intersect(hm_sample, sample_id_subgroup2)], c(1, 2), mean, na.rm = TRUE)
hist_mat_list_subgroup2[[k]] = copyAttr(mat, hist_mat_list_subgroup2[[k]])
hist_mat_list_diff[[k]] = hist_mat_list_subgroup1[[k]] - hist_mat_list_subgroup2[[k]]
}
save(meth_mat, meth_mat_1, meth_mat_2, meth_mat_diff, cor_mat_list,
hist_mat_list, hist_mat_list_subgroup1, hist_mat_list_subgroup2, hist_mat_list_diff, file = rdata_file)
}
add_boxplot_of_cgi_length = function(ht_list) {
gl = cgi_width
anno_name = "cgi_width"
row_order_list = row_order(ht_list)
lt = lapply(row_order_list, function(ind) gl[ind])
bx = boxplot(lt, plot = FALSE)$stats
n = length(row_order_list)
x_ind = (seq_len(n) - 0.5)/n
w = 1/n*0.5
decorate_annotation(anno_name, slice = 1, {
rg = range(bx)
rg[1] = rg[1] - (rg[2] - rg[1])*0.1
rg[2] = rg[2] + (rg[2] - rg[1])*0.1
pushViewport(viewport(y = unit(1, "npc") + unit(1, "mm"), just = "bottom", height = unit(2, "cm"), yscale = rg))
grid.rect(gp = gpar(col = "black"))
grid.segments(x_ind - w/2, bx[5, ], x_ind + w/2, bx[5, ], default.units = "native", gp = gpar(lty = 1:2))
grid.segments(x_ind - w/2, bx[1, ], x_ind + w/2, bx[1, ], default.units = "native", gp = gpar(lty = 1:2))
grid.segments(x_ind, bx[1, ], x_ind, bx[5, ], default.units = "native", gp = gpar(lty = 1:2))
grid.rect(x_ind, colMeans(bx[c(4, 2), ]), width = w, height = bx[4, ] - bx[2, ], default.units = "native", gp = gpar(fill = "white", lty = 1:2))
grid.segments(x_ind - w/2, bx[3, ], x_ind + w/2, bx[3, ], default.units = "native", gp = gpar(lty = 1:2))
grid.yaxis(main = FALSE, gp = gpar(fontsize = 8))
grid.text(anno_name, y = unit(1, "npc") + unit(2.5, "mm"), gp = gpar(fontsize = 14), just = "bottom")
upViewport()
})
}
expr = EXPR[cgi2$gene_id, sample_id, drop = FALSE]
## calculate row orders
expr_mean = rowMeans(expr[, SAMPLE[sample_id, ]$subgroup == "subgroup1"]) -
rowMeans(expr[, SAMPLE[sample_id, ]$subgroup == "subgroup2"])
expr_split = ifelse(expr_mean > 0, "high", "low")
expr_split = factor(expr_split, levels = c("high", "low"))
set.seed(123)
target_index = attr(meth_mat, "target_index")
meth_split = kmeans(meth_mat[, target_index], centers = 2)$cluster
x = tapply(rowMeans(meth_mat[, target_index]), meth_split, mean)
od = structure(order(x), names = names(x))
meth_split = paste0("cluster", od[as.character(meth_split)])
combined_split = paste(meth_split, expr_split, sep = "|")
merge_row_order = function(l_list) {
do.call("c", lapply(l_list, function(l) {
if(sum(l) == 0) return(integer(0))
if(sum(l) == 1) return(which(l))
dend1 = as.dendrogram(hclust(dist_by_closeness2(mat[l, ])))
dend1 = reorder(dend1, rowMeans(mat[l, ]))
od = order.dendrogram(dend1)
which(l)[od]
}))
}
row_order = merge_row_order(list(
combined_split == "cluster1|high" & km == 1,
combined_split == "cluster1|high" & km == 2,
combined_split == "cluster1|high" & km == 3,
combined_split == "cluster1|high" & km == 4,
combined_split == "cluster1|low" & km == 1,
combined_split == "cluster1|low" & km == 2,
combined_split == "cluster1|low" & km == 3,
combined_split == "cluster1|low" & km == 4,
combined_split == "cluster2|high" & km == 1,
combined_split == "cluster2|high" & km == 2,
combined_split == "cluster2|high" & km == 3,
combined_split == "cluster2|high" & km == 4,
combined_split == "cluster2|low" & km == 1,
combined_split == "cluster2|low" & km == 2,
combined_split == "cluster2|low" & km == 3,
combined_split == "cluster2|low" & km == 4
))
n_heatmap = 0
col = c("-1" = "darkgreen", "0" = "white", "1" = "red")
## heatmap for expression
# columns are clustered for each subgroup
dend1 = as.dendrogram(hclust(dist(t(expr[, SAMPLE$subgroup == "subgroup1"]))))
hc1 = as.hclust(reorder(dend1, colMeans(expr[, SAMPLE$subgroup == "subgroup1"])))
expr_col_od1 = hc1$order
dend2 = as.dendrogram(hclust(dist(t(expr[, SAMPLE$subgroup == "subgroup2"]))))
hc2 = as.hclust(reorder(dend2, colMeans(expr[, SAMPLE$subgroup == "subgroup2"])))
expr_col_od2 = hc2$order
expr_col_od = c(which(SAMPLE$subgroup == "subgroup1")[expr_col_od1], which(SAMPLE$subgroup == "subgroup2")[expr_col_od2])
cor_col_fun = colorRamp2(c(-1, 0, 1), c("darkgreen", "white", "red"))
axis_name = c("-5kb", "Start", "End", "5kb")
ht_list = Heatmap(expr, name = "expr", show_row_names = FALSE,
show_column_names = FALSE, width = unit(5, "cm"), show_column_dend = FALSE, cluster_columns = FALSE, column_order = expr_col_od,
top_annotation = HeatmapAnnotation(group = SAMPLE[sample_id, ]$group, sample_type = SAMPLE[sample_id, ]$sample_type, subgroup = SAMPLE[sample_id, ]$subgroup,
col = list(group = COLOR$group, sample_type = COLOR$sample_type, subgroup = COLOR$subgroup), show_annotation_name = TRUE, annotation_name_side = "left"),
column_title = "Expression", show_row_dend = FALSE,
use_raster = TRUE, raster_quality = 2)
gap = unit(0.3, "cm")
n_heatmap = n_heatmap + 1
ht_list = ht_list + Heatmap(km[names(cgi2)], name = "km_groups", col = km_col, show_row_names = FALSE,
width = unit(1, "cm"))
gap = unit.c(gap, unit(1, "cm"))
mat_mix = mat
mat_mix[mat == 0] = mat_vice[mat == 0]
ht_list = ht_list +
EnrichedHeatmap(mat_mix, name = qq("@{type}CR"), col = col, split = strd,
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(neg_col = "darkgreen", pos_col = "red", lty = 1:2))),
top_annotation_height = unit(2, "cm"), column_title = qq("sig@{type}CR"),
use_raster = TRUE, raster_quality = 2, combined_name_fun = NULL, axis_name = axis_name)+
rowAnnotation(cgi_width = row_anno_points(cgi_width, axis = TRUE, gp = gpar(col = "#00000040")),
width = unit(1, "cm"))
n_heatmap = n_heatmap + 1
gap = unit(c(0.3, 1, 1), "cm")
ht_list = ht_list + EnrichedHeatmap(mat_corr, col = cor_col_fun, name = qq("correlation"),
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(pos_col = "red", neg_col = "darkgreen", lty = 1:2))),
top_annotation_height = unit(2, "cm"), column_title = qq("corr_meth"), axis_name = axis_name,
use_raster = TRUE, raster_quality = 2)
gap = unit.c(gap, unit(1, "cm"))
n_heatmap = n_heatmap + 1
# methylation
ht_list = ht_list + EnrichedHeatmap(meth_mat, col = colorRamp2(c(0, 0.5, 1), c("blue", "white", "red")),
name = "methylation", column_title = qq("meth"), axis_name = axis_name,
heatmap_legend_param = list(title = "methylation"),
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(col = "red", lty = 1:2))),
use_raster = TRUE, raster_quality = 2)
gap = unit.c(gap, unit(1, "cm"))
n_heatmap = n_heatmap + 1
generate_diff_color_fun = function(x) {
q = quantile(x, c(0.05, 0.95))
max_q = max(abs(q))
colorRamp2(c(-max_q, 0, max_q), c("#3794bf", "#FFFFFF", "#df8640"))
}
ht_list = ht_list + EnrichedHeatmap(meth_mat_diff, col = generate_diff_color_fun(meth_mat_diff),
name = "methylation_diff", column_title = qq("meth_diff"), axis_name = axis_name,
heatmap_legend_param = list(title = "methylation_diff"),
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(pos_col = "#df8640", neg_col = "#3794bf", lty = 1:2))),
use_raster = TRUE, raster_quality = 2)
gap = unit.c(gap, unit(1, "cm"))
n_heatmap = n_heatmap + 1
ht_list2 = NULL
ht_list1 = NULL
# correlation to histone marks
for(i in seq_along(cor_mat_list)) {
if(i == 3) {
ht_list1 = ht_list
ht_list = NULL
}
if(length(cor_mat_list[[i]]) > 1) {
anno_line_col = ifelse(mean(cor_mat_list[[i]], na.rm = TRUE) > 0, "red", "darkgreen")
ht_list = ht_list + EnrichedHeatmap(cor_mat_list[[i]], col = cor_col_fun, name = qq("corr_@{MARKS[i]}"),
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(pos_col = "red", neg_col = "darkgreen", lty = 1:2))),
top_annotation_height = unit(2, "cm"), column_title = qq("corr_@{MARKS[i]}"), axis_name = axis_name,
use_raster = TRUE, raster_quality = 2)
gap = unit.c(gap, unit(1, "cm"))
n_heatmap = n_heatmap + 1
}
ht_list = ht_list + EnrichedHeatmap(hist_mat_list[[i]], col = colorRamp2(quantile(hist_mat_list[[i]], c(0, 0.95)), c("white", "purple")), name = qq("@{MARKS[i]}_1"),
column_title = qq("@{MARKS[i]}"), axis_name = axis_name,
heatmap_legend_param = list(title = qq("@{MARKS[i]}_density")),
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(col = "purple", lty = 1:2))),
use_raster = TRUE, raster_quality = 2)
gap = unit.c(gap, unit(1, "cm"))
n_heatmap = n_heatmap + 1
ht_list = ht_list + EnrichedHeatmap(hist_mat_list_diff[[i]], name = qq("@{MARKS[i]}_diff"), col = generate_diff_color_fun(hist_mat_list_diff[[i]]),
column_title = qq("@{MARKS[i]}_diff"), axis_name = axis_name,
heatmap_legend_param = list(title = qq("@{MARKS[i]}_diff")),
top_annotation = HeatmapAnnotation(lines1 = anno_enriched(gp = gpar(pos_col = "#df8640", neg_col = "#3794bf", lty = 1:2))),
use_raster = TRUE, raster_quality = 2)
gap = unit.c(gap, unit(1, "cm"))
n_heatmap = n_heatmap + 1
}
ht_list2 = ht_list
lines_lgd = Legend(at = c("high", "low"), title = "Lines", legend_gp = gpar(lty = 1:2), type = "lines")
# following chunk is necessary because Legend() needs to open a new graphic device
for(i in seq_along(dev.list())) {
dev.off()
}
pdf(qq("@{OUTPUT_DIR}/plots/cr_enrichedheatmap_cgi_by_gene_@{type}_fdr_@{cutoff}_methdiff_@{meandiff}.pdf"), width = n_heatmap + 4, height = 10)
ht_list = Heatmap(expr_split, show_row_names = FALSE, name = "expr_split", col = c("high" = "red", "low" = "darkgreen"), width = unit(5, "mm")) + ht_list1
foo = draw(ht_list, annotation_legend_list = list(lines_lgd),
cluster_rows = FALSE, row_order = row_order, show_row_dend = FALSE,
column_title = qq("cluster by methylation, @{nrow(mat)} rows"), split = meth_split, row_sub_title_side = "left",
show_heatmap_legend = FALSE, annotation_legend_side = "right")
add_boxplot_of_cgi_length(foo)
i = 0
for(f in names(ht_list1@ht_list)) {
if(grepl("expr|annotation|CGI|km|CR", f)) next
decorate_column_title(f, {
grid.rect(height = unit(0.8, "npc"), gp = gpar(fill = brewer.pal(8, "Set2")[as.integer(i/3)+1], col = NA))
grid.text(ht_list1@ht_list[[f]]@column_title, gp = gpar(fontsize = 14))
})
i = i + 1
}
ht_list = Heatmap(expr_split, show_row_names = FALSE, name = "expr_split", col = c("high" = "red", "low" = "darkgreen"), width = unit(5, "mm")) + ht_list2
foo = draw(ht_list, annotation_legend_list = list(lines_lgd),
cluster_rows = FALSE, row_order = row_order, show_row_dend = FALSE,
column_title = qq("cluster by methylation, @{nrow(mat)} rows"), split = meth_split, row_sub_title_side = "left",
show_heatmap_legend = FALSE)
for(f in names(ht_list2@ht_list)) {
decorate_column_title(f, {
grid.rect(height = unit(0.8, "npc"), gp = gpar(fill = brewer.pal(8, "Set2")[as.integer(i/3)+1], col = NA))
grid.text(ht_list2@ht_list[[f]]@column_title, gp = gpar(fontsize = 14))
})
i = i + 1
}
dev.off()
# for(cutoff in c(0.1, 0.05, 0.01)) {
# for(meandiff in c(0, 0.1, 0.2, 0.3)) {
# for(type in c("pos", "neg")) {
# cmd = qq("Rscript-3.1.2 /icgc/dkfzlsdf/analysis/B080/guz/roadmap_analysis/re_analysis/scripts/04.correlated_regions_enrichedheatmap_at_cgi.R --rerun --type @{type} --cutoff @{cutoff} --meandiff @{meandiff}")
# cmd = qq("perl /home/guz/project/development/ngspipeline2/qsub_single_line.pl '-l walltime=20:00:00,mem=15G -N cr_enrichedheatmap_at_cgi_@{type}_fdr_@{cutoff}_methdiff_@{meandiff}' '@{cmd}'")
# system(cmd)
# }
# }
# }
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.