-
Notifications
You must be signed in to change notification settings - Fork 0
/
copykat_pipeline.R
402 lines (348 loc) · 20.9 KB
/
copykat_pipeline.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
#
# Plain run
# Rscript copykat_pipeline.R \
# -c gene_counts.csv.gz \
# -g chromosome_coordinates_hg19.txt \
# -o output \
#
# Optional parameters
# -u umap_with_celltype_and_cluster.txt
# -l cell_labels.txt
#
library(optparse)
library(copykat)
library(tidyr)
library(gtools)
library(ggplot2)
library(cowplot)
#' Attempts to call a chromosome arm as altered by comparing the data distribution along the arm of previously determined anuploid cells with that of
#' diploid cells via a Gaussian distribution. A cell is called as having the chromosome arm altered when more than "calling_threshold" proportion of
#' data bins fall outside of "pvalue_signif_cutoff" of the Gaussian distribution formed by the cells marked as diploid
#'
#' @param chrom The chromosome that is being evaluated
#' @param arm Either "p" or "q" denominating the arm to evaluate
#' @param CNA_data Data frame with the "CNAmat" item from the Copykat output
#' @param genome_coords Data frame with genomic coordinates identifying chromosome and centromere start/end. Required columns: chr, start, cen.left.base, cen.right.base, end
#' @param cell_labels Data frame with calls wether a cell is diploid or aneuploid taken from Copykat output list item "prediction" with two columns: cell.names, copykat.pred
#' @param calling_threshold The proportion of bins to be significant to call a chromosome arm as altered
#' @param pvalue_signif_cutoff How far a bin must be in the tail of a Gaussian distribution to call it significant
#'
assess_chromarm = function(chrom, arm, CNA_data, genome_coords, cell_labels, calling_threshold=0.5, pvalue_signif_cutoff=0.05) {
cent_start = genome_coords$cen.left.base[genome_coords$chr==chrom]
cent_end = genome_coords$cen.right.base[genome_coords$chr==chrom]
if (arm=="p") {
chromname = paste0(chrom, "p")
chrom_aneuploid = CNA_data[CNA_data$chrom==chrom & CNA_data$chrompos < cent_start, cell_labels$cell.names[cell_labels$copykat.pred=="aneuploid"]]
chrom_diploid = CNA_data[CNA_data$chrom==chrom & CNA_data$chrompos < cent_start, cell_labels$cell.names[cell_labels$copykat.pred=="diploid"]]
} else {
chromname = paste0(chrom, "q")
chrom_aneuploid = CNA_data[CNA_data$chrom==chrom & CNA_data$chrompos > cent_end, cell_labels$cell.names[cell_labels$copykat.pred=="aneuploid"]]
chrom_diploid = CNA_data[CNA_data$chrom==chrom & CNA_data$chrompos > cent_end, cell_labels$cell.names[cell_labels$copykat.pred=="diploid"]]
}
chrom_aneuploid_m = tidyr::pivot_longer(chrom_aneuploid, cols=colnames(chrom_aneuploid))
chrom_diploid_m = tidyr::pivot_longer(chrom_diploid, cols=colnames(chrom_diploid))
chrom_diploid_mean = mean(chrom_diploid_m$value)
chrom_diploid_sd = sd(chrom_diploid_m$value)
# If there is no data we should not continue
if (nrow(chrom_diploid_m)==0 | nrow(chrom_aneuploid_m)==0) {
num_aneuploid = sum(cell_labels$copykat.pred=="aneuploid")
return(list(chromname=chromname, plot=NULL, pvalues=rep(NA, num_aneuploid), is_signif=rep(NA, num_aneuploid), chrom_aneuploid=chrom_aneuploid, chrom_diploid=chrom_diploid, cells_with_alteration=NULL, calling_threshold=calling_threshold))
}
pvalues = matrix(NA, nrow=nrow(chrom_aneuploid), ncol=ncol(chrom_aneuploid))
is_signif = matrix(NA, nrow=ncol(chrom_aneuploid), ncol=1)
for (i in 1:ncol(chrom_aneuploid)) {
pvalues[,i] = dnorm(chrom_aneuploid[,i], mean=chrom_diploid_mean, sd=chrom_diploid_sd)
is_signif[i] = (sum(pvalues[,i] < pvalue_signif_cutoff) / length(pvalues[,i]))
}
cells_with_alteration = colnames(chrom_aneuploid)[is_signif > calling_threshold]
chrom_aneuploid_m$altered = chrom_aneuploid_m$name %in% cells_with_alteration
df <- data.frame(PF = chrom_diploid_m$value)
p1 = ggplot(df) + aes(x = PF) +
geom_histogram(mapping=aes(y =..density..),
# breaks = seq(-50, 50, by = 10),
colour = "grey",
fill = "white",
bins=30) +
stat_function(fun = dnorm, args = list(mean = chrom_diploid_mean, sd = chrom_diploid_sd), colour="blue", size=1.5) +
xlim(-0.75, 0.75) + theme_cowplot() + ggtitle(paste0("Normal cells - chrom ", chromname)) + xlab("Relative expression") + ylab("Density")
df2 = data.frame(PF=chrom_aneuploid_m$value, Altered=factor(chrom_aneuploid_m$altered, levels=c(TRUE, FALSE)))
p2 = ggplot() +
geom_histogram(data=df2,
mapping=aes(x=PF, y =..density.., fill=Altered),
# breaks = seq(-50, 50, by = 10),
colour = "black",
position="stack",
bins=30) +
scale_fill_manual(values = c("red", "white")) +
stat_function(fun = dnorm, args = list(mean = chrom_diploid_mean, sd = chrom_diploid_sd), colour="blue", size=1.5) +
xlim(-0.75, 0.75) + theme_cowplot() + ggtitle(paste0("Aneuploid cells - chrom ", chromname)) + xlab("Relative expression") + ylab("Density")
plot_combined = plot_grid(p1, p2, ncol=2, rel_widths=c(8.5/20, 11.5/20))
return(list(chromname=chromname, plot=plot_combined, pvalues=pvalues, is_signif=is_signif, chrom_aneuploid=chrom_aneuploid, chrom_diploid=chrom_diploid, cells_with_alteration=cells_with_alteration, calling_threshold=calling_threshold))
}
#' Takes Copykat normalised and binned expression values and attempts to definitevely call a chromosome arm as altered
#'
#' @param CNA_data Data frame with the "CNAmat" item from the Copykat output
#' @param predictions Data frame with calls wether a cell is diploid or aneuploid taken from Copykat output list item "prediction" with two columns: cell.names, copykat.pred
#' @param genome_coords Data frame with genomic coordinates identifying chromosome and centromere start/end. Required columns: chr, start, cen.left.base, cen.right.base, end
#' @param calling_threshold Fraction of bins that need to be significantly different from the normals for a chrom arm to be called altered (Default: 0.5)
#'
call_alterations = function(CNA_data, predictions, genome_coords, calling_threshold=0.5) {
calls = matrix(NA, nrow=23*2, ncol=sum(predictions$copykat.pred=="aneuploid"))
calls_res = list()
chrom_names = unique(CNA_data$chrom)
for (i in 1:length(chrom_names)) {
chrom = chrom_names[i]
res_q = assess_chromarm(chrom = chrom, arm="q", CNA_data = CNA_data, genome_coords = genome_coords, cell_labels=predictions, calling_threshold=calling_threshold)
res_p = assess_chromarm(chrom = chrom, arm="p", CNA_data = CNA_data, genome_coords = genome_coords, cell_labels=predictions, calling_threshold=calling_threshold)
calls[(i*2)-1,] = res_p$is_signif
calls[(i*2),] = res_q$is_signif
calls_res[[paste0(chrom, "p")]] = res_p
calls_res[[paste0(chrom, "q")]] = res_q
}
colnames(calls) = predictions$cell.names[predictions$copykat.pred=="aneuploid"]
rownames(calls) = unlist(lapply(unique(CNA_data$chrom), function(x) paste0(x, c("p", "q"))))
return(calls_res)
}
theme_plot = function(p) {
p = p + theme(axis.text = element_text(colour="black",size=10,face="plain"),
axis.title = element_text(colour="black",size=12,face="plain"),
strip.text.x = element_text(colour="black",size=12,face="plain", angle=90),
strip.text.y = element_text(colour="black",size=12,face="plain", angle=360),
strip.background = element_blank(),
panel.spacing.x = unit(1, "mm"),
panel.spacing.y = unit(3, "mm"),
plot.title = element_text(colour="black",size=12,face="plain",hjust = 0.5))
return(p)
}
#' Plot the distributions used to call chromosomes as altered
#'
#' @param umap_coords Data frame with UMAP coordinates, this file must also have a column named "cluster"
#' @param calls_res List output from "call_alterations"
#' @param chrom_names Vector with chromosome names to plot
#' @param outfileprefix Prefix of the output figures, including a possible directory
#'
plot_chrom_dists = function(umap_coords, calls_res, chrom_names, outfileprefix) {
for (chrom in chrom_names) {
p_arm = paste0(chrom, "p")
q_arm = paste0(chrom, "q")
p_altered_cells = calls_res[[p_arm]]$cells_with_alteration
q_altered_cells = calls_res[[q_arm]]$cells_with_alteration
if (!is.null(umap_coords)) {
umap_coords$Altered = factor(umap_coords$cell %in% unique(c(p_altered_cells, q_altered_cells)), levels=c(FALSE, TRUE))
umap_coords = umap_coords[order(umap_coords$Altered),]
p = ggplot(umap_coords) + aes(x=UMAP_1, y=UMAP_2, colour=Altered) + geom_point(size=1.5) +
scale_colour_manual(values = c("grey", "red")) +
ggtitle(paste0("Altered cells chromosome ", chrom)) + theme_cowplot() + theme(legend.position="bottom")
} else {
p = NULL
}
png(file.path(outfileprefix, paste0("chr", chrom, ".png")), height=500, width=1500)
print(plot_grid(
plot_grid(calls_res[[p_arm]]$plot, calls_res[[q_arm]]$plot, nrow=2),
p,
ncol=2))
dev.off()
}
}
#' Make a pileup plot
#'
#' @param dat Data frame with columns: chrom, chrompos, sumpositive, sumnegative
#' @param ylabel Y axis label to be used in the plot
#' @param plot_title Title of the plot
#'
plot_pileup = function(dat, ylabel, plot_title) {
p = ggplot(dat) +
geom_point(mapping=aes(x=chrompos, y=sumpositive), colour="red", size=0.75) + geom_line(mapping=aes(x=chrompos, y=sumpositive), colour="red") +
geom_point(mapping=aes(x=chrompos, y=sumnegative), colour="blue", size=0.75) + geom_line(mapping=aes(x=chrompos, y=sumnegative), colour="blue") +
facet_grid(~chrom, space="free", scale="free") +
geom_hline(yintercept=0, colour="black") +
geom_vline(data=data.frame(f=2, x=1), mapping=aes(xintercept=x), linetype=2, size=0.5, col="black") +
xlab("Chromosome / Position") + ylab(ylabel) +
ggtitle(plot_title) +
theme_bw() +
theme(panel.spacing = unit(0, "mm"),
panel.border = element_blank(),
panel.background = element_blank(),
panel.grid = element_blank(),
strip.background = element_blank(),
strip.text.x = element_text(colour="black",size=20,face="plain", angle=90),
axis.text.x=element_blank(),
axis.text.y = element_text(colour="black",size=20,face="plain"),
axis.ticks.x = element_blank(),
axis.line.x=element_blank(),
axis.title = element_text(colour="black",size=20,face="plain"),
plot.title = element_text(colour="black",size=22,face="plain")) #,hjust = 0.5
return(p)
}
plot_pileup_all_cells = function(calls_res, CNA_data, outfileprefix) {
dummy_coords = data.frame(cell=colnames(CNA_data)[4:ncol(CNA_data)], All_Cells="")
plot_pileup_by_annotation(dummy_coords, calls_res, CNA_data, outfileprefix, annotation="All_Cells")
}
#' Make a pileup plot per cluster that shows the proportion of cells with an alteration per genomic coordinate
#'
#' @param umap_coords Data frame with UMAP coordinates, this file must also have a column named "cluster"
#' @param calls_res List output from "call_alterations"
#' @param CNA_data Data frame with the "CNAmat" item from the Copykat output
#' @param outfileprefix Prefix of the output figures, including a possible directory
#'
plot_cluster_pileup = function(umap_coords, calls_res, CNA_data, outfileprefix) {
plot_pileup_by_annotation(umap_coords, calls_res, CNA_data, outfileprefix, annotation="cluster")
}
#' Make a pileup plot per celltype that shows the proportion of cells with an alteration per genomic coordinate
#'
#' @param umap_coords Data frame with UMAP coordinates, this file must also have a column named "cluster"
#' @param calls_res List output from "call_alterations"
#' @param CNA_data Data frame with the "CNAmat" item from the Copykat output
#' @param outfileprefix Prefix of the output figures, including a possible directory
#'
plot_celltype_pileup = function(umap_coords, calls_res, CNA_data, outfileprefix) {
plot_pileup_by_annotation(umap_coords, calls_res, CNA_data, outfileprefix, annotation="celltype")
}
#' Make a pileup plot per annotation column that shows the number or proportion of cells with an alteration per genomic coordinate
#'
#' @param umap_coords Data frame with UMAP coordinates, this file must also have a column named "cluster"
#' @param calls_res List output from "call_alterations"
#' @param CNA_data Data frame with the "CNAmat" item from the Copykat output
#' @param outfileprefix Prefix of the output figures, including a possible directory
#' @param annotation The column name in the "umap_coords" data frame to be used to group cells. A pair of plots is made for each group
#'
plot_pileup_by_annotation = function(umap_coords, calls_res, CNA_data, outfileprefix, annotation) {
for (anno in sort(unique(umap_coords[,annotation]))) {
cells_cluster = umap_coords$cell[umap_coords[,annotation]==anno]
cells_cluster = cells_cluster[cells_cluster %in% colnames(CNA_data)]
dat_cluster = CNA_data[, c("chrom", "chrompos", "abspos", cells_cluster)]
dat_cluster$chrom = factor(dat_cluster$chrom, levels=gtools::mixedsort(unique(dat_cluster$chrom)))
dat_cluster$sumpositive = apply(dat_cluster[, 4:ncol(dat_cluster)], 1, function(x) sum(x[x>0]))
dat_cluster$sumnegative = apply(dat_cluster[, 4:ncol(dat_cluster)], 1, function(x) sum(x[x<0]))
p = plot_pileup(dat_cluster, "Summed signal", paste0("Summed signal ", gsub("_", " ", annotation), " ", anno))
png(file.path(outfileprefix, paste0("summedpileup_", anno, ".png")), height=400, width=1000)
print(p)
dev.off()
num_cells_cluster = ncol(dat_cluster)-3
for (chrom in unique(dat_cluster$chrom)) {
cells_altered_chrom_p = calls_res[[paste0(chrom, "p")]]$cells_with_alteration
chrom_start = genome_coords$start[genome_coords$chr==chrom]
chrom_end = genome_coords$cen.left.base[genome_coords$chr==chrom]
selected_bins = dat_cluster$chrom==chrom & dat_cluster$chrompos >= chrom_start & dat_cluster$chrompos < chrom_end
selected_cells = cells_altered_chrom_p[cells_altered_chrom_p %in% colnames(dat_cluster)]
# round to -1 or 1 for cells_altered_chrom_p - depending on whether the value is >0 or <0
dat_cluster[selected_bins, selected_cells] = apply(dat_cluster[selected_bins, selected_cells, drop=F], 2, function(x) {
x_out = x
x_out[x < 0] = floor(x[x < 0])
x_out[x > 0] = ceiling(x[x > 0])
x_out
})
# set to 0 for anything else
not_selected_cells = colnames(dat_cluster)
not_selected_cells = not_selected_cells[!not_selected_cells %in% c(selected_cells, "chrom", "chrompos", "abspos")]
dat_cluster[selected_bins, not_selected_cells] = 0
cells_altered_chrom_q = calls_res[[paste0(chrom, "q")]]$cells_with_alteration
chrom_start = genome_coords$cen.right.base[genome_coords$chr==chrom]
chrom_end = genome_coords$end[genome_coords$chr==chrom]
selected_bins = dat_cluster$chrom==chrom & dat_cluster$chrompos >= chrom_start & dat_cluster$chrompos < chrom_end
selected_cells = cells_altered_chrom_q[cells_altered_chrom_q %in% colnames(dat_cluster)]
# round to -1 or 1 for cells_altered_chrom_q - depending on whether the value is >0 or <0
dat_cluster[selected_bins, selected_cells] = apply(dat_cluster[selected_bins, selected_cells, drop=F], 2, function(x) {
x_out = x
x_out[x < 0] = floor(x[x < 0])
x_out[x > 0] = ceiling(x[x > 0])
x_out
})
# set to 0 for anything else
not_selected_cells = colnames(dat_cluster)
not_selected_cells = not_selected_cells[!not_selected_cells %in% c(selected_cells, "chrom", "chrompos", "abspos")]
dat_cluster[selected_bins, not_selected_cells] = 0
# outside of the selected bins, set everything to 0
bins_outside_coords = dat_cluster$chrompos < genome_coords$start[genome_coords$chr==chrom] |
dat_cluster$chrompos > genome_coords$end[genome_coords$chr==chrom] |
(dat_cluster$chrompos > genome_coords$cen.left.base[genome_coords$chr==chrom] & dat_cluster$chrompos < genome_coords$cen.right.base[genome_coords$chr==chrom])
dat_cluster[dat_cluster$chrom==chrom & bins_outside_coords, 4:ncol(dat_cluster)] = 0
}
dat_cluster$sumpositive = apply(dat_cluster[, 4:ncol(dat_cluster)], 1, function(x) sum(x[x>0]))
dat_cluster$sumnegative = apply(dat_cluster[, 4:ncol(dat_cluster)], 1, function(x) sum(x[x<0]))
# divide sums by total cells
dat_cluster$sumpositive = dat_cluster$sumpositive / length(cells_cluster)
dat_cluster$sumnegative = dat_cluster$sumnegative / length(cells_cluster)
# plot
p = plot_pileup(dat_cluster, "Fraction of cells called", paste0("Fraction of cells with alterations in ", gsub("_", " ", annotation), " ", anno, " N=", num_cells_cluster))
p = p + ylim(-1, 1)
png(file.path(outfileprefix, paste0("fraccellspileup_", anno, ".png")), height=400, width=1000)
print(p)
dev.off()
}
}
option_list = list(
make_option(c("-c", "--counts_matrix"), type="character", default=NULL, help="File containing the raw counts matrix", metavar="character"),
make_option(c("-u", "--umap_coords"), type="character", default=NULL, help="Optional file containing UMAP coordinates and optionally columns called `cluster` and `celltype`", metavar="character"),
make_option(c("-g", "--genome_coords"), type="character", default=NULL, help="File with chromosome and centromere start/end coordinates", metavar="character"),
make_option(c("-o", "--output_dir"), type="character", default=NULL, help="File with chromosome and centromere start/end coordinates", metavar="character"),
make_option(c("-l", "--cell_labels"), type="character", default=NULL, help="Optional file with labels per cell whether it is aneuploid or diploid", metavar="character"),
make_option(c("-p", "--num_cores"), type="integer", default=4, help="Number of cores to use", metavar="character")
)
opt_parser = OptionParser(option_list=option_list)
opt = parse_args(opt_parser)
counts_file = opt$counts_matrix
umap_file = opt$umap_coords
genome_coords_file = opt$genome_coords
outputdir = opt$output_dir
cell_labels_file = opt$cell_labels
num_cores = opt$num_cores
calling_threshold = 0.1 # The proportion of bins to be significant to call a chromosome arm as altered
if (is.null(counts_file)) { stop("Counts matrix file must be supplied") }
if (is.null(genome_coords_file)) { stop("Genome coords file must be supplied") }
if (is.null(outputdir)) { stop("Output dir must be supplied") }
print("Reading in data")
counts = read.table(counts_file, header=T, stringsAsFactors=F)
if (!is.null(umap_file)) {
umap_coords = read.table(umap_file, header=T, stringsAsFactors=F)
} else {
umap_coords = NULL
}
genome_coords = read.table(genome_coords_file, header=T)
print("Phase 1: Running Copykat")
dir.create(file.path(outputdir, "01_copykat"), showWarnings=F)
if (is.null(cell_labels_file)) {
norm.cell.names = NULL
} else {
cell_labels = read.table(cell_labels_file, header=T, stringsAsFactors=F)
norm.cell.names = cell_labels$cell.names[cell_labels$copykat.pred=="diploid"]
}
copykat_res <- copykat(rawmat=counts,
id.type="S",
cell.line="no",
ngene.chr=5,
win.size=25,
norm.cell.names=norm.cell.names,
KS.cut=0.2,
sam.name=file.path(outputdir, "01_copykat", "copykat"),
distance="euclidean",
n.cores=num_cores)
# copykat_res = readRDS(file.path(outputdir, "01_copykat", "copykat_fullresult.rds"))
saveRDS(file=file.path(outputdir, "01_copykat", "copykat_fullresult.rds"), copykat_res)
predictions <- data.frame(copykat_res$prediction)
print("Phase 2: Calling CNAs")
dir.create(file.path(outputdir, "02_calling"), showWarnings=F)
CNA_data = data.frame(copykat_res$CNAmat)
calls_res = call_alterations(CNA_data, predictions, genome_coords, calling_threshold)
plot_chrom_dists(umap_coords, calls_res, unique(CNA_data$chrom), file.path(outputdir, "02_calling"))
# Summarise the calls as TRUE/FALSE per cell per chromosome arm
output = as.data.frame(matrix(FALSE, ncol=length(calls_res)+1, nrow=ncol(counts)))
colnames(output) = c("cell", names(calls_res))
output$cell = colnames(counts)
for (chromarm in names(calls_res)) {
output[output$cell %in% calls_res[[chromarm]]$cells_with_alteration, chromarm] = TRUE
}
write.table(output, file=file.path(outputdir, "called_alterations.txt"), quote=F, sep="\t", row.names=F)
# Make a plot showing the overall alterations called
plot_pileup_all_cells(calls_res, CNA_data, outputdir)
if (!is.null(umap_coords) && "cluster" %in% colnames(umap_coords)) {
print("Phase 3: Plotting pileup per cluster")
dir.create(file.path(outputdir, "03_plot_per_cluster"), showWarnings=F)
plot_cluster_pileup(umap_coords, calls_res, CNA_data, file.path(outputdir, "03_plot_per_cluster"))
}
if (!is.null(umap_coords) && "celltype" %in% colnames(umap_coords)) {
print("Phase 3: Plotting pileup per celltype")
dir.create(file.path(outputdir, "03_plot_per_celltype"), showWarnings=F)
plot_celltype_pileup(umap_coords, calls_res, CNA_data, file.path(outputdir, "03_plot_per_celltype"))
}
print("Done")