-
Notifications
You must be signed in to change notification settings - Fork 2
/
clean_solid.R
executable file
·247 lines (200 loc) · 8.15 KB
/
clean_solid.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
library(readxl)
solid_xlsx = "ProXe_NIBRX_models_annotations_20190329_sk_xl2011.xlsx"
## read in solid metadata as solid_meta
solid_meta <- readxl::read_excel(
file.path(data_outside_app_dir, solid_xlsx),
sheet = "Header_Data"
)
solid <- readxl::read_excel(
file.path(data_outside_app_dir, solid_xlsx),
sheet = 1
)
match_vector = c("MODELID" = "PDX_Name",
"HISTOLOGY"="COSMIC_Type",
"PRIMARY_SITE"="COSMIC_Primary_Site",
"HIST_SUBTYPE1"="COSMIC_Subtype")
# swap solid_meta$Internal_Column_Header names and colnames(solid)
for (name in names(match_vector)) {
if (name %in% colnames(solid)){
colnames(solid)[colnames(solid) == name] <- match_vector[name]
}
if (name %in% solid_meta$Internal_Column_Header) {
solid_meta$Internal_Column_Header[solid_meta$Internal_Column_Header == name] <- match_vector[name]
}
}
# change sample prefix
solid$PDX_Name <- gsub("^", "NIBR-", solid$PDX_Name)
# add availability column manually
solid$Distribution_Permissions <- "academic, industry-sponsored academic, and industry"
##### ---- RNA-seq data manipulation for plotting ------ #####
# TODO: manipulate this, create a graph out of it,
# and also fold it in (link too?) with 'solid' df above through a merge.
gao_rna <- readxl::read_excel(
file.path(data_outside_app_dir, "Gao_et_al_PDXs_suppl_table_Nat_Med_2015.xlsx"),
sheet = "RNAseq_fpkm"
)
gao_rna <- as.data.frame(gao_rna)
rownames(gao_rna) <- gao_rna$Sample
gao_rna$Sample <- NULL
library(pheatmap)
# change sample prefix
colnames(gao_rna) <- gsub("X-", "NIBR-", colnames(gao_rna))
# remove genes that are all zero
rows_to_keep <- apply(gao_rna, 1, function(row) {
sum(row) > 0
})
gao_rna <- gao_rna[rows_to_keep, ]
# # keep only genes with rpkm > 5 in > 20 samples
# rows_to_keep <- apply(gao_rna,1,function(row){
# sum(row > 5) > 20
# })
# gr_filt <- gao_rna[rows_to_keep,]
#
# # calculate variance/mean ratio
# if(anyNA(gr_filt)) stop("NAs in gao_rna filtered")
# var_mean_ratio_sort <- sort(
# apply(gr_filt,1,function(row){
# var(row) / mean(row)
# })
# , decreasing = FALSE)
# create a log2-scaled gr
# grfl <- log2(gr_filt + 0.01)
# create z-scores = (value - mean) / sqrt(variance)
# grfz <- t(scale(t(gr_filt)))
# take top X by var/mean ratio
# X = 500
# grfz_topX <- grfz[names(var_mean_ratio_sort)[1:X],]
# # filter for one type
# grfz_topX_type = grfz_topX[,
# solid[na.omit(solid$COSMIC_Primary_Site=="pancreas" & solid$COSMIC_Type=="carcinoma"),]$PDX_Name
# ]
# plot
# breaksList = seq(floor(min(grfz)), -floor(min(grfz)),length.out=length(my_palette))
# pheatmap::pheatmap(grfz_topX_type,
# color = my_palette,
# breaks = breaksList
# )
## TODO:
# add heatmap plot to app.
# plot metadata on the side.
# maybe make interactive
# look at stats
# sums are different, thus these are probably FPKM/RPKM
# range(colSums(gao_rna)) # 260k - 905k
# compute a mean of all samples for comparison.
# min(gao_rna[gao_rna > 0]) # 0.01 across all samples
# gao_rna_pc <- gao_rna + 1
# gao_meds <- apply(gao_rna_pc,1,median)
# gao_fc <- sweep(gao_rna_pc,MARGIN = 1,STATS = gao_meds, FUN = "/")
# grp_lfc <- log2(gao_fc)
# # determine number of genes >2^6-fold up
# up <- apply(grp_lfc,2,function(col){
# sum(col > 6)
# })
# grp <- data.frame( Sample = names(up), num_genes_64x_up = up, stringsAsFactors = F)
# # which genes
# grp$genes_64x_up = apply(grp_lfc,2,function(col){
# paste0(names(col[col>6]),collapse=" | ")
# })
# # determine number of genes <2^5-fold down
# grp$num_genes_16x_dn <- apply(grp_lfc,2,function(col){
# sum(col < -4)
# })
# # which genes
# grp$genes_16x_dn = apply(grp_lfc,2,function(col){
# paste0(names(col[col < -4]),collapse=" | ")
# })
# Ideas for more columns
# 1. take top and bottom 10 genes, show
# 2. take genes above threshold that cross-reference with oncogene list (already in app)
# 3. Compute most commonly-changed genes
#
# Ideas for plots:
# 1. PCA
# 2. correlation heatmaps.
if (F) {
# try PCA
# ok to use raw rpkm values?
gpca <- t(gao_rna)
# remove constant cols
gpca <- gpca[, apply(gpca, 2, var) != 0]
gpca_prcomp <- prcomp(gpca, scale = TRUE) # note sure whether I'm supposed to be transposing. Think this is right.
plot(gpca_prcomp)
summary(gpca_prcomp)
biplot(gpca_prcomp)
# gpca_princomp <- princomp(t(gpca)) # just playing, could be wrong. failed.
# biplot(gpca_princomp)
}
####### --------- mutation data ---------- #############
## read in mutation data ##
gao_mut <- readxl::read_excel(
file.path(data_outside_app_dir, "Gao_et_al_PDXs_suppl_table_Nat_Med_2015.xlsx"),
sheet = "pdxe_mut_and_cn2"
)
# change sample prefix
gao_mut$Sample <- gsub("X-", "NIBR-", gao_mut$Sample)
gao_mut <- gao_mut[grepl("Mut", x = gao_mut$Category), ]
gao_mut$Alteration <- gsub(",.*", "", gao_mut$Details)
gao_mut$Pct <- as.character(paste0(as.numeric(gsub(".*,", "", gao_mut$Details)) * 100, "%"))
gao_mut$Details2 <- paste(gao_mut$Gene, gao_mut$Alteration, "in", gao_mut$Pct, "of reads")
# Separate Known/Likely vs VUS
gao_pos <- gao_mut[gao_mut$Category %in% c("MutKnownFunctional", "MutLikelyFunctional"), ]
gao_vus <- gao_mut[gao_mut$Category == "MutNovel", ]
# For each, create summary df
unique_samples <- levels(as.factor(gao_pos$Sample))
outdf <- data.frame(Sample = unique_samples, Genes = rep(NA, length(unique_samples)), Details = rep(NA, length(unique_samples)))
for (sam in unique_samples) {
gao_sam <- gao_pos[gao_pos$Sample == sam, ]
outdf[outdf$Sample == sam, "Genes"] <- paste(gao_sam$Gene, collapse = " | ")
outdf[outdf$Sample == sam, "Details"] <- paste(gao_sam$Details2, collapse = " | ")
outdf[outdf$Sample == sam, "Number"] <- nrow(gao_sam)
}
names(outdf) <- c("Sample", "PDX_Mutations_Positive", "PDX_Mutations_Details", "PDX_Mutations_Count")
unique_samples <- levels(as.factor(gao_vus$Sample))
outdf2 <- data.frame(Sample = unique_samples, Genes = rep(NA, length(unique_samples)), Details = rep(NA, length(unique_samples)))
for (sam in unique_samples) {
gao_sam <- gao_vus[gao_vus$Sample == sam, ]
outdf2[outdf2$Sample == sam, "Genes"] <- paste(gao_sam$Gene, collapse = " | ")
outdf2[outdf2$Sample == sam, "Details"] <- paste(gao_sam$Details2, collapse = " | ")
}
names(outdf2) <- c("Sample", "PDX_VUS", "PDX_VUS_Details")
gao_muts <- merge(outdf, outdf2, by = "Sample")
rm(list = c("gao_mut", "gao_pos", "gao_sam", "gao_vus", "outdf", "outdf2"))
names(gao_muts)[1] <- "PDX_Name"
# merge new columns back to 'solid' to show in app
# solid <- merge(solid,grp,by="Sample")
solid <- merge(solid, gao_muts, by = "PDX_Name")
rm(gao_muts)
##-- Switch to PRoXe_Column_Header -- ##
# convert column names from PRIMAGRAFTS name to desired PRoXe name
# order 'meta' by 'meta$Interal_Column_Header' matching names(df)
solid_meta <- solid_meta[match(names(solid), solid_meta$Internal_Column_Header), ]
if (!all(names(solid) == solid_meta$Internal_Column_Header)) stop("ordering incorrect")
names(solid) <- solid_meta$PRoXe_Column_Header
## -- choose which columns should be visible, invisible, etc. -- ##
# levels(as.factor(meta$Visible_Invisible)) # [1] "cond_vis" "delete" "ob_invis" "ob_vis"
# 1. get rid of 'delete'
to_delete <- which(solid_meta$Visible_Invisible == "delete")
if (length(to_delete) > 0) {
solid <- solid[, -to_delete]
solid_meta2 <- solid_meta[-to_delete, ]
} else {
solid_meta2 <- solid_meta
}
# 2. sort all columns by categories
solid_meta2$Visible_Invisible_int <- rep(NA_integer_, nrow(solid_meta2))
solid_meta2[solid_meta2$Visible_Invisible == "ob_vis", ]$Visible_Invisible_int <- 1
solid_meta2[solid_meta2$Visible_Invisible == "cond_vis", ]$Visible_Invisible_int <- 2
if (nrow(solid_meta2[solid_meta2$Visible_Invisible == "ob_invis", ]) > 0) {
solid_meta2[solid_meta2$Visible_Invisible == "ob_invis", ]$Visible_Invisible_int <- 3
}
library(plyr)
solid_meta2 <- arrange(solid_meta2, Visible_Invisible_int)
solid <- solid[, solid_meta2$PRoXe_Column_Header]
# 3. store values for demarcation
categ_count <- table(solid_meta2$Visible_Invisible)
condVis_ind_solid <- unname(categ_count["ob_vis"] + 1) # marks beginning of cond vis
obInvisRet_ind_solid <- unname(condVis_ind + categ_count["cond_vis"]) # marks beginning of invis but retained
## change appropriate columns to type ##
solid <- as.data.frame(solid)
solid <- convert.magic(solid, solid_meta2$Data_Type)