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
|
####################################################################
##### Testing of numerical output
##### Very difficult to automate, because of stochasticity in methods.
## with current data set used, CBS is +/- OK, but not the HMM-based methods.
##### To obtain the RData for the comparisons and run this. Of course,
##### be sure parameters to segment are the right ones (and have the
##### same values as the web run).
#### Beware: for CBS, if by Array*Chromosome, the pre-smoothing stage is not identical!!
#### Thus, code is changed in web server, to ensure usage of A method.
#### Prepare data set
dataIn <- read.table("two.sample.shuffled.num.test",
sep = "\t", header = TRUE,
comment.char = "")
######### Turn chromosome into a number
tmpchr <- sub("chr", "", dataIn$Chromosome)
Chrom <- as.numeric(as.character(tmpchr))
Chrom[tmpchr == "X"] <- 23
## Verify all is OK
table(Chrom, dataIn$Chromosome)
########### Order data
mid.point <- dataIn$UG.Start +
0.5 * (dataIn$UG.End - dataIn$UG.Start)
orderindex <- order(Chrom, mid.point,
dataIn$UG.Start,
dataIn$UG.End)
## Verify no duplicate spots. Not a big deal, but better if not.
tmp <- paste(Chrom, mid.point, sep = ".")
length(tmp) == length(unique(tmp))
Chrom.ord <- Chrom[orderindex]
dataCGH <- as.matrix(dataIn[orderindex, 5:6])
Pos.ord <- mid.point[orderindex]
make.hmm.obj <- function(data, chrom, pos) {
Clone <- 1:nrow(data)
obj <- create.aCGH(data.frame(data),
data.frame(Clone = Clone,
Chrom = chrom,
kb = pos))
return(obj)
}
doHMM <- function(data, nsamp) {
res <- find.hmm.states(data, aic = TRUE, bic = FALSE)
hmm(data) <- res
r1 <<- hmm(data)
m <- matrix(NA, nrow = nrow(data$hmm$states.hmm[[1]]),
ncol = nsamp)
for(j in 1:nsamp) {
m[, j] <- mergeLevels(data$hmm$states.hmm[[1]][, 2 + (6 * j)],
data$hmm$states.hmm[[1]][, 2 + (6 * j) - 2])$vecMerged
}
write.table(file = "HMM.R.output.txt", m)
return(m)
}
fHMM <- function(data, chrom, pos) {
obj <- make.hmm.obj(data, chrom, pos)
return(doHMM(obj, ncol(data)))
}
fBioHMM <- function(data, Chrom, Pos) {
uchrom <- unique(Chrom)
m <- matrix(NA, nrow = nrow(data), ncol = ncol(data))
for (j in 1:ncol(data)) {
dataj <- data[, j]
smoothed <- NULL
for (ic in uchrom) {
ydat <- dataj[Chrom == ic]
n <- length(ydat)
res <- snapCGH:::fit.model(sample = 1, chrom = ic, dat = matrix(ydat, ncol = 1),
datainfo = data.frame(Name = 1:n, Chrom = rep(ic, n),
Position = Pos[Chrom == ic]))
smoothed <- c(smoothed, res$out.list$mean)
}
m[, j] <- mergeLevels(data[, j], smoothed)$vecMerged
}
write.table(file = "BioHMM.R.output.txt", m)
return(m)
}
fGLAD <- function(data, Chrom, Pos) {
m <- matrix(NA, nrow = nrow(data), ncol = 2 * ncol(data))
Pos <- if (is.null(Pos)) (1:nrow(data)) else Pos
for(j in 1:ncol(data)) {
x <- data[, j]
tmpf <- data.frame(LogRatio = x,
PosOrder = Pos,
Chromosome = Chrom)
tmpf <- list(profileValues = tmpf)
class(tmpf) <- "profileCGH"
outglad <- glad.profileCGH(tmpf)
m[, ((2 * j) - 1)] <- outglad$profileValues$Smoothing
m[, (2 * j)] <- outglad$profileValues$ZoneGNL
}
write.table(file = "GLAD.R.output.txt", m)
return(m)
}
make.cbs.obj <- function(tmp, chrom, pos) {
cna.obj <- CNA(tmp, chrom, pos,
data.type = "logratio")
return(cna.obj)
}
mergeDNAcopy <- function(object) {
numarrays <- ncol(object$data) - 2
## zz: we must get numarray from object
if(!(inherits(object, "DNAcopy")))
stop("This function can only be applied to DNAcopy objects")
merged_segments <- list()
merged_segments$chrom.numeric <- object$data$chrom ## verify its numeric zz!!
merged_segments$segm <- list()
for(arraynum in 1:numarrays) {
obs <- object$data[, 2 + arraynum]
segmented <-
object$output[object$output$ID ==
colnames(object$data)[2 + arraynum], ]
segmentus <- object$data$maploc
for(i in 1:nrow(segmented)) {
segmentus[(segmented[i,'loc.end'] >= segmentus) &
(segmented[i,'loc.start'] <= segmentus)] <-
segmented[i,'seg.mean']
}
segmentus2 <- mergeLevels(obs, segmentus)$vecMerged
classes.ref <- which.min(abs(unique(segmentus2)))
classes.ref <- unique(segmentus2)[classes.ref]
ref <- rep(0, length(segmentus2))
ref[segmentus2 > classes.ref] <- 1
ref[segmentus2 < classes.ref] <- -1
merged_segments$segm[[arraynum]] <- cbind(obs,
merged.mean = segmentus2,
alteration = ref)
}
class(merged_segments) <- c(class(merged_segments),
"mergedDNAcopy")
return(merged_segments)
}
fCBSo <- function(data, chrom, pos) {
## this is the old one, with prunning
data <- make.cbs.obj(data, chrom, pos)
smoothed <- smooth.CNA(data)
segmented <- segment(smoothed, undo.splits = "prune", nperm = 10000)
res <- mergeDNAcopy(segmented)
return(res)
}
fCBS <- function(data, chrom, pos) {
data <- make.cbs.obj(data, chrom, pos)
smoothed <- smooth.CNA(data)
segmented <- segment(smoothed, undo.splits = "none", nperm = 10000)
res <- mergeDNAcopy(segmented)
write.table(file = "CBS.R.output.txt", res)
return(res)
}
fc3 <- function(x, y, tol = 1e-06) {
tmp <- max(abs(x - y))
if(tmp < tol) {
return("OK")
} else{
return("Failed? ---but look at figures; tests can seem to fail when really OK")
}
}
cbs.compare <- function(res.cbs.r, res.web, filename = "cbscompare.pdf") {
rcbs <- matrix(unlist(res.cbs.r$segm), ncol = 3 * length(res.cbs.r))
res.web <- res.web[ , -(1:6)]
ncl <- ncol(res.web)
pdf(file = filename)
for(i in seq(from = 2, to = ncl - 1, by = 3)) {
par(pty = "s")
plot(jitter(res.web[, i]), jitter(rcbs[, i]))
abline(a = 0, b = 1, lty = 2)
}
dev.off()
fc3(rcbs, res.web, tol = 5e-05) ## cbs only provides four digits
}
hmm.compare <- function(rd, wd, filename = "hmmcompare.pdf") {
ncrd <- ncol(rd)
wdo <- wd[, 8 + (0:(ncrd - 1)) * 3]
pdf(file = filename)
for(i in 1:ncrd) {
par(pty = "s")
plot(jitter(rd[, i]), jitter(wdo[, i]))
abline(a = 0, b = 1, lty = 2)
}
dev.off()
fc3(rd, wdo)
}
glad.compare <- function(rd, wd) {
ncrd <- ncol(rd)/2
nocols <- 1:6
nocols <- c(nocols, 7 + (0:(ncrd - 1))* 3)
wdo <- wd[, -nocols]
pdf(file = "gladcompare.pdf")
for(i in 1:ncrd) {
par(pty = "s")
plot(jitter(rd[, i]), jitter(wdo[, i]))
abline(a = 0, b = 1, lty = 2)
}
dev.off()
fc3(rd, wdo)
}
test.cbs <- function() {
## this will not work if tilingArray loaded ...
try(detach("package:tilingArray"))
library(DNAcopy)
library(aCGH)
cbs.r.out <- fCBS(dataCGH, Chrom.ord, pos = 1:length(Chrom.ord))
cbs.web.out <- read.table("CBS.web.output.txt", sep = "\t", header = TRUE)
rv <- cbs.compare(cbs.r.out, cbs.web.out)
save(file = "cbs.test.RData", list = ls())
return(rv)
}
test.hmm <- function() {
library(aCGH)
library(snapCGH)
hmm.r.out <- fHMM(dataCGH, Chrom.ord, pos = 1:length(Chrom.ord))
hmm.web.out <- read.table("HMM.web.output.txt", sep = "\t", header = TRUE)
rv <- hmm.compare(hmm.r.out, hmm.web.out)
save(file = "hmm.test.RData", list = ls())
return(rv)
}
test.biohmm <- function() {
library(aCGH)
library(snapCGH)
biohmm.r.out <- fBioHMM(dataCGH, Chrom.ord, Pos.ord)
biohmm.web.out <- read.table("BioHMM.web.output.txt", sep = "\t", header = TRUE)
rv <- hmm.compare(biohmm.r.out, biohmm.web.out, filename = "biohmmcompare.pdf")
save(file = "biohmm.test.RData", list = ls())
return(rv)
}
test.glad <- function() {
library(GLAD)
glad.r.out <- fGLAD(dataCGH, Chrom.ord, NULL)
glad.web.out <- read.table("GLAD.web.output.txt", sep = "\t", header = TRUE)
rv <- glad.compare(glad.r.out, glad.web.out)
save(file = "glad.test.RData", list = ls())
return(rv)
}
## Because of name conflicts, cannot load all packs. at same time
## library(DNAcopy)
## library(aCGH)
## ## Note that there is variability among runs, which can lead to apparently
## ## failed tests. The current data set seems to (almost) always lead to the same
## ## solutions, but it need not be so.
## ## the actual computations and comparisons
## cbs.r.out <- fCBS(dataCGH, Chrom.ord, pos = 1:length(Chrom.ord))
## cbs.web.out <- read.table("CBS.output.txt", sep = "\t", header = TRUE)
## cbs.compare(cbs.r.out, cbs.web.out)
## ## If you think there is variability among runs of CBS, wait till you see HMMs!
## library(snapCGH)
## hmm.r.out <- fHMM(dataCGH, Chrom.ord, pos = 1:length(Chrom.ord))
## hmm.web.out <- read.table("HMM.output.txt", sep = "\t", header = TRUE)
## hmm.compare(hmm.r.out, hmm.web.out)
## biohmm.r.out <- fBioHMM(dataCGH, Chrom.ord, Pos.ord)
## biohmm.web.out <- read.table("BioHMM.output.txt", sep = "\t", header = TRUE)
## hmm.compare(biohmm.r.out, biohmm.web.out)
## ## If we wanted to get really fancy, we'd try to formally compare results
## ## from multiple runs of each. But the plots above are enough, probably.
## ## GLAD seems fairly stable.
## glad.r.out <- fGLAD(dataCGH, Chrom.ord, NULL)
## glad.web.out <- read.table("GLAD.output.txt", sep = "\t", header = TRUE)
## glad.compare(glad.r.out, glad.web.out)
|