Efrain Gonzalez / Cleaning and Fixing Data with R

Download as
Browse Code ยป

Commit 2677909e39ee690329745832c97a3fb94154909c

Authored by Efrain Gonzalez
1 parent cb1063ceb9
Exists in master

Added "Gene symbol" to glossary for geneIDNam

Showing 1 changed file with 8 additions and 8 deletions   Show diff stats
1 1
2 #Efrain H. Gonzalez 2 #Efrain H. Gonzalez
3 #6/22/2017 3 #6/22/2017
4 options(digits = 11) 4 options(digits = 11)
5 #Libraries required to run the code 5 #Libraries required to run the code
6 library(pryr) 6 library(pryr)
7 library(MASS) 7 library(MASS)
8 library(dplyr) 8 library(dplyr)
9 library(tidyr) 9 library(tidyr)
10 library(readr) 10 library(readr)
11 library(stringr) 11 library(stringr)
12 12
13 13
14 #Necessary Functions 14 #Necessary Functions
15 #1#Function for handling the changing of row names and column names 15 #1#Function for handling the changing of row names and column names
16 chngrownm <- function(mat){ 16 chngrownm <- function(mat){
17 row <- dim(mat)[1] 17 row <- dim(mat)[1]
18 col <- dim(mat)[2] 18 col <- dim(mat)[2]
19 e <- 1 19 e <- 1
20 r <- 1 20 r <- 1
21 a <- 1 21 a <- 1
22 h <- 1 22 h <- 1
23 g <- 1 23 g <- 1
24 o <- 1 24 o <- 1
25 for(e in 1:col){ 25 for(e in 1:col){
26 if("!Sample_source_name_ch1"==mat[1,e]){ 26 if("!Sample_source_name_ch1"==mat[1,e]){
27 colnames(mat)[e] <- "Brain_Region" 27 colnames(mat)[e] <- "Brain_Region"
28 } else if("!Sample_title" == mat[1,e]){ 28 } else if("!Sample_title" == mat[1,e]){
29 colnames(mat)[e] <- "Title" 29 colnames(mat)[e] <- "Title"
30 } else if("!Sample_geo_accession" == mat[1,e]){ 30 } else if("!Sample_geo_accession" == mat[1,e]){
31 colnames(mat)[e] <- "ID_REF" 31 colnames(mat)[e] <- "ID_REF"
32 } else{ 32 } else{
33 if(grepl("Sex|gender|Gender|sex",mat[2,e])==TRUE){ 33 if(grepl("Sex|gender|Gender|sex",mat[2,e])==TRUE){
34 colnames(mat)[e] <- paste0("Sex",r) 34 colnames(mat)[e] <- paste0("Sex",r)
35 r = r + 1 35 r = r + 1
36 } 36 }
37 if(grepl("postmorteminterval|PMI|pmi|interval",mat[2,e])==TRUE){ 37 if(grepl("postmorteminterval|PMI|pmi|interval",mat[2,e])==TRUE){
38 colnames(mat)[e] <- paste0("PMI",a) 38 colnames(mat)[e] <- paste0("PMI",a)
39 a = a + 1 39 a = a + 1
40 } 40 }
41 if(grepl("age|Age|AGE",mat[2,e])==TRUE){ 41 if(grepl("age|Age|AGE",mat[2,e])==TRUE){
42 colnames(mat)[e] <- paste0("Age",h) 42 colnames(mat)[e] <- paste0("Age",h)
43 h = h + 1 43 h = h + 1
44 } 44 }
45 if(grepl("braak|b&b",mat[2,e])==TRUE){ 45 if(grepl("braak|b&b",mat[2,e])==TRUE){
46 colnames(mat)[e] <- paste0("Braak",g) 46 colnames(mat)[e] <- paste0("Braak",g)
47 g = g + 1 47 g = g + 1
48 } 48 }
49 if(grepl("group|disease|control|AD|normal|diagnosis|Alzheimer|Control|Normal",mat[2,e])==TRUE){ 49 if(grepl("group|disease|control|AD|normal|diagnosis|Alzheimer|Control|Normal",mat[2,e])==TRUE){
50 colnames(mat)[e] <- paste0("Group",o) 50 colnames(mat)[e] <- paste0("Group",o)
51 o = o + 1 51 o = o + 1
52 } 52 }
53 53
54 } 54 }
55 e = e + 1 55 e = e + 1
56 } 56 }
57 mat 57 mat
58 } 58 }
59 59
60 #2#Function for reorganizing information within the columns 60 #2#Function for reorganizing information within the columns
61 cinfo <- function(mat){ 61 cinfo <- function(mat){
62 col <- dim(mat)[2] 62 col <- dim(mat)[2]
63 j <-2 63 j <-2
64 for(j in 2:col){ 64 for(j in 2:col){
65 if(grepl("Group",colnames(mat)[j]) == TRUE){ 65 if(grepl("Group",colnames(mat)[j]) == TRUE){
66 mat[,j] <- gsub(".+:\\s|\\s.+;.+","",mat[,j]) 66 mat[,j] <- gsub(".+:\\s|\\s.+;.+","",mat[,j])
67 } else if(grepl("Age",colnames(mat)[j])==TRUE){ 67 } else if(grepl("Age",colnames(mat)[j])==TRUE){
68 mat[,j] <- gsub("\\D","",mat[,j])%>% 68 mat[,j] <- gsub("\\D","",mat[,j])%>%
69 as.integer() 69 as.integer()
70 } else if(grepl("Sex",colnames(mat)[j])==TRUE){ 70 } else if(grepl("Sex",colnames(mat)[j])==TRUE){
71 mat[,j] <- gsub(".+:\\s","",mat[,j]) 71 mat[,j] <- gsub(".+:\\s","",mat[,j])
72 } else if(grepl("PMI",colnames(mat)[j])==TRUE){ 72 } else if(grepl("PMI",colnames(mat)[j])==TRUE){
73 mat[,j] <- gsub("[^0-9\\.]","",mat[,j])%>% 73 mat[,j] <- gsub("[^0-9\\.]","",mat[,j])%>%
74 as.numeric() 74 as.numeric()
75 } else if(grepl("Braak",colnames(mat)[j])==TRUE){ 75 } else if(grepl("Braak",colnames(mat)[j])==TRUE){
76 mat[,j]<-gsub(".+:\\s","",mat[,j])%>% 76 mat[,j]<-gsub(".+:\\s","",mat[,j])%>%
77 as.roman()%>% 77 as.roman()%>%
78 as.integer() 78 as.integer()
79 } 79 }
80 j=j+1 80 j=j+1
81 } 81 }
82 mat 82 mat
83 } 83 }
84 84
85 #3#Function for labeling the gene IDs without names 85 #3#Function for labeling the gene IDs without names
86 NAFIXING <- function(GIDNAM){ 86 NAFIXING <- function(GIDNAM){
87 row <- dim(GIDNAM)[1] 87 row <- dim(GIDNAM)[1]
88 i <- 1 88 i <- 1
89 for(i in 1:row){ 89 for(i in 1:row){
90 if(grepl("^NA\\s*$",GIDNAM[i,2])==TRUE||is.na(GIDNAM[i,2])==TRUE){ 90 if(grepl("^NA\\s*$",GIDNAM[i,2])==TRUE||is.na(GIDNAM[i,2])==TRUE){
91 GIDNAM[i,2] <- GIDNAM[i,1] 91 GIDNAM[i,2] <- GIDNAM[i,1]
92 } 92 }
93 i <- i + 1 93 i <- i + 1
94 } 94 }
95 GIDNAM 95 GIDNAM
96 } 96 }
97 97
98 #4#Function for changing the gene ID to gene name 98 #4#Function for changing the gene ID to gene name
99 cgeneID <- function(GeneName,DATA){ 99 cgeneID <- function(GeneName,DATA){
100 nj <- t(GeneName) 100 nj <- t(GeneName)
101 nq <- t(DATA) 101 nq <- t(DATA)
102 colGene <- dim(nj)[2] 102 colGene <- dim(nj)[2]
103 colDATA <- dim(nq)[2] 103 colDATA <- dim(nq)[2]
104 j <- 1 104 j <- 1
105 for(j in 1:colDATA){ 105 for(j in 1:colDATA){
106 #where is that gene id located within the GPL file 106 #where is that gene id located within the GPL file
107 chngreq <- grep(paste0("^",nq[1,j],"$"),nj[1,]) 107 chngreq <- grep(paste0("^",nq[1,j],"$"),nj[1,])
108 if(is.na(sum(chngreq))==FALSE){ 108 if(is.na(sum(chngreq))==FALSE){
109 if(sum(chngreq) > 0){ 109 if(sum(chngreq) > 0){
110 nq[1,j] <- gsub(paste0("^",nq[1,j],"$"),nj[2,chngreq],nq[1,j]) 110 nq[1,j] <- gsub(paste0("^",nq[1,j],"$"),nj[2,chngreq],nq[1,j])
111 } 111 }
112 } 112 }
113 j <- j + 1 113 j <- j + 1
114 } 114 }
115 nq 115 nq
116 } 116 }
117 #cgeneID <- function(GeneName,DATA){ 117 #cgeneID <- function(GeneName,DATA){
118 # colGene <- dim(GeneName)[2] 118 # colGene <- dim(GeneName)[2]
119 # j <- 1 119 # j <- 1
120 # for(j in 1:colGene){ 120 # for(j in 1:colGene){
121 # chngsreq <- grep(paste0("^",GeneName[1,j],"$"),DATA[1,]) 121 # chngsreq <- grep(paste0("^",GeneName[1,j],"$"),DATA[1,])
122 # if(is.na(sum(chngsreq))==FALSE){ 122 # if(is.na(sum(chngsreq))==FALSE){
123 # if(sum(chngsreq) > 0){ 123 # if(sum(chngsreq) > 0){
124 # DATA[1,chngsreq] <- gsub(paste0("^",GeneName[1,j]),GeneName[2,j],DATA[1,chngsreq]) 124 # DATA[1,chngsreq] <- gsub(paste0("^",GeneName[1,j]),GeneName[2,j],DATA[1,chngsreq])
125 # } 125 # }
126 # } 126 # }
127 # j = j+1 127 # j = j+1
128 # } 128 # }
129 # DATA 129 # DATA
130 #} 130 #}
131 131
132 #5#Function for adjusting the gene names 132 #5#Function for adjusting the gene names
133 gcnames <- function(DiData,usecol=1){ 133 gcnames <- function(DiData,usecol=1){
134 nuruns <- dim(DiData)[2] 134 nuruns <- dim(DiData)[2]
135 i = 1 135 i = 1
136 nwnam <- rep("0",length.out=nuruns) 136 nwnam <- rep("0",length.out=nuruns)
137 for(i in 1:nuruns){ 137 for(i in 1:nuruns){
138 if(length(strsplit(colnames(DiData)[i],"///|//")[[1]]) >= usecol){ 138 if(length(strsplit(colnames(DiData)[i],"///|//")[[1]]) >= usecol){
139 nwnam[i]=str_trim(strsplit(colnames(DiData)[i],"///|//")[[1]][usecol]) 139 nwnam[i]=str_trim(strsplit(colnames(DiData)[i],"///|//")[[1]][usecol])
140 } else{ 140 } else{
141 nwnam[i]=str_trim(strsplit(colnames(DiData)[i],"///|//")[[1]][1]) 141 nwnam[i]=str_trim(strsplit(colnames(DiData)[i],"///|//")[[1]][1])
142 } 142 }
143 143
144 } 144 }
145 nwnam 145 nwnam
146 146
147 } 147 }
148 148
149 #6# Function for discretizing the data 149 #6# Function for discretizing the data
150 dndat <- function(NDATA){ 150 dndat <- function(NDATA){
151 rownd <- dim(NDATA)[1] 151 rownd <- dim(NDATA)[1]
152 colnd <- dim(NDATA)[2] 152 colnd <- dim(NDATA)[2]
153 DDATA <- matrix(0,nrow=rownd,ncol=colnd) 153 DDATA <- matrix(0,nrow=rownd,ncol=colnd)
154 colnames(DDATA) <- colnames(NDATA) 154 colnames(DDATA) <- colnames(NDATA)
155 i <- 1 155 i <- 1
156 for(i in 1:rownd){ 156 for(i in 1:rownd){
157 j <- 1 157 j <- 1
158 for(j in 1:colnd){ 158 for(j in 1:colnd){
159 if(is.na(NDATA[i,j])==FALSE){ 159 if(is.na(NDATA[i,j])==FALSE){
160 160
161 if(NDATA[i,j] < -1){ 161 if(NDATA[i,j] < -1){
162 DDATA[i,j]=0L 162 DDATA[i,j]=0L
163 } else if(NDATA[i,j] > 1){ 163 } else if(NDATA[i,j] > 1){
164 DDATA[i,j]=2L 164 DDATA[i,j]=2L
165 } else if(-1 <= NDATA[i,j] && NDATA[i,j] <= 1){ 165 } else if(-1 <= NDATA[i,j] && NDATA[i,j] <= 1){
166 DDATA[i,j]=1L 166 DDATA[i,j]=1L
167 } 167 }
168 } else{ 168 } else{
169 DDATA[i,j] = NDATA[i,j] 169 DDATA[i,j] = NDATA[i,j]
170 } 170 }
171 j = j + 1 171 j = j + 1
172 } 172 }
173 i = i + 1 173 i = i + 1
174 } 174 }
175 DDATA 175 DDATA
176 } 176 }
177 177
178 178
179 #MajorFunction#This is the function that does everything else 179 #MajorFunction#This is the function that does everything else
180 THEFT <- function(){ 180 THEFT <- function(){
181 #Set working directory based on the directory of the series matrix file Currently only works for windows 181 #Set working directory based on the directory of the series matrix file Currently only works for windows
182 wd <- getwd() 182 wd <- getwd()
183 #list.files() 183 #list.files()
184 #gsub("wd",wd,"Do you want to clean all data files in the directory wd?") 184 #gsub("wd",wd,"Do you want to clean all data files in the directory wd?")
185 numDAT <- switch(EXPR = menu(choices = c("Yes","No"),title = gsub("wd",wd,"Do you want to clean all data files in the directory wd?")) + 1,cat("Nothing done\n"),1L,2L) 185 numDAT <- switch(EXPR = menu(choices = c("Yes","No"),title = gsub("wd",wd,"Do you want to clean all data files in the directory wd?")) + 1,cat("Nothing done\n"),1L,2L)
186 GSEfileloc <- grep("^GSE.+\\.txt\\.gz$",list.files()) 186 GSEfileloc <- grep("^GSE.+\\.txt\\.gz$",list.files())
187 GSEfloc <- list.files()[GSEfileloc] 187 GSEfloc <- list.files()[GSEfileloc]
188 #ALL DATA FILES WILL BE CLEANED 188 #ALL DATA FILES WILL BE CLEANED
189 if(numDAT == 1){ 189 if(numDAT == 1){
190 #indexing the data files 190 #indexing the data files
191 n <- 1 191 n <- 1
192 for(n in 1: length(GSEfloc)){ 192 for(n in 1: length(GSEfloc)){
193 alz <- GSEfloc[n] 193 alz <- GSEfloc[n]
194 194
195 #Working with the wordy part of the document 195 #Working with the wordy part of the document
196 alzword <- alz %>% 196 alzword <- alz %>%
197 read_delim(delim ="\t",comment = "!Series",col_names = FALSE)%>% 197 read_delim(delim ="\t",comment = "!Series",col_names = FALSE)%>%
198 filter(grepl("!Sample",X1))%>% 198 filter(grepl("!Sample",X1))%>%
199 filter(!grepl("!Sample_contact",X1)) 199 filter(!grepl("!Sample_contact",X1))
200 200
201 #Getting the GPL file 201 #Getting the GPL file
202 genena <- grep("_platform_id",alzword$X1) %>% 202 genena <- grep("_platform_id",alzword$X1) %>%
203 alzword$X2[.] %>% 203 alzword$X2[.] %>%
204 str_trim(.) %>% 204 str_trim(.) %>%
205 paste0("^",.,"\\D") %>% 205 paste0("^",.,"\\D") %>%
206 grep(.,list.files()) %>% 206 grep(.,list.files()) %>%
207 list.files()[.] 207 list.files()[.]
208 208
209 #Find out if it is a soft GPL file or not 209 #Find out if it is a soft GPL file or not
210 soft <- strsplit(genena,"[\\|/]") %>% 210 soft <- strsplit(genena,"[\\|/]") %>%
211 .[[1]] %>% 211 .[[1]] %>%
212 .[length(.)] %>% 212 .[length(.)] %>%
213 grepl("soft",.) 213 grepl("soft",.)
214 214
215 ##Changing row names and column names: 215 ##Changing row names and column names:
216 ALZWORD <- t(alzword) 216 ALZWORD <- t(alzword)
217 rownames(ALZWORD)=NULL 217 rownames(ALZWORD)=NULL
218 colnames(ALZWORD) <- colnames(ALZWORD,do.NULL=FALSE) 218 colnames(ALZWORD) <- colnames(ALZWORD,do.NULL=FALSE)
219 ALZWORD <- chngrownm(ALZWORD)[-1,] 219 ALZWORD <- chngrownm(ALZWORD)[-1,]
220 ALZWORD <- ALZWORD%>% 220 ALZWORD <- ALZWORD%>%
221 as.data.frame(.,stringsAsFactors = FALSE)%>% 221 as.data.frame(.,stringsAsFactors = FALSE)%>%
222 dplyr::select(-starts_with("col")) 222 dplyr::select(-starts_with("col"))
223 223
224 ##Reorganizing information within the columns and final clinical data 224 ##Reorganizing information within the columns and final clinical data
225 ALZWORDF <- cinfo(ALZWORD) 225 ALZWORDF <- cinfo(ALZWORD)
226 226
227 227
228 #Working with Actual Data part of file 228 #Working with Actual Data part of file
229 alzdat <- alz %>% 229 alzdat <- alz %>%
230 read_delim(delim="\t",col_names=TRUE,comment = "!",skip=1) 230 read_delim(delim="\t",col_names=TRUE,comment = "!",skip=1)
231 ALZDAT <- t(alzdat[,-1]) 231 ALZDAT <- t(alzdat[,-1])
232 rownames(ALZDAT)=NULL 232 rownames(ALZDAT)=NULL
233 233
234 ##Is there a clean version of the GPL file available? 234 ##Is there a clean version of the GPL file available?
235 gplnum <- strsplit(genena,"[\\|/]") %>% 235 gplnum <- strsplit(genena,"[\\|/]") %>%
236 .[[1]] %>% 236 .[[1]] %>%
237 .[length(.)] %>% 237 .[length(.)] %>%
238 gsub("\\D","",.) 238 gsub("\\D","",.)
239 clfileex <- sum(grepl(paste0("Clean_GPL",gplnum),list.files())) 239 clfileex <- sum(grepl(paste0("Clean_GPL",gplnum),list.files()))
240 if(clfileex >= 1){ 240 if(clfileex >= 1){
241 #use the clean version 241 #use the clean version
242 geneIDNam <- paste0("Clean_GPL",gplnum,".txt") %>% 242 geneIDNam <- paste0("Clean_GPL",gplnum,".txt") %>%
243 read_delim(delim="\t",col_names = c("ID","Symbol"), comment = "!") 243 read_delim(delim="\t",col_names = c("ID","Symbol"), comment = "!")
244 244
245 } else if(clfileex == 0){ 245 } else if(clfileex == 0){
246 ##Lets Create a clean version 246 ##Lets Create a clean version
247 247
248 ##Gene ID to Gene Name 248 ##Gene ID to Gene Name
249 if(soft == TRUE){ 249 if(soft == TRUE){
250 #Check to see if there is already a file containing information on soft files 250 #Check to see if there is already a file containing information on soft files
251 fileex <- sum(grepl("GPL_ID_LOC.txt",list.files())) 251 fileex <- sum(grepl("GPL_ID_LOC.txt",list.files()))
252 if(fileex == 1){ 252 if(fileex == 1){
253 #Check to see if this GPL soft file has been used before 253 #Check to see if this GPL soft file has been used before
254 IDF <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>% 254 IDF <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>%
255 .$GPL_FILE_NUM%>% 255 .$GPL_FILE_NUM%>%
256 grepl(gplnum,.) %>% 256 grepl(gplnum,.) %>%
257 sum() 257 sum()
258 if(IDF == 1){ 258 if(IDF == 1){
259 IDLOCAL <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>% 259 IDLOCAL <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>%
260 .$GPL_FILE_NUM%>% 260 .$GPL_FILE_NUM%>%
261 grep(gplnum,.) 261 grep(gplnum,.)
262 idlocgpl <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>% 262 idlocgpl <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>%
263 .$LOC_ID %>% 263 .$LOC_ID %>%
264 .[IDLOCAL] 264 .[IDLOCAL]
265 geneIDNam <- genena %>% 265 geneIDNam <- genena %>%
266 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idlocgpl) %>% 266 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idlocgpl) %>%
267 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 267 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
268 } else if(IDF == 0){ 268 } else if(IDF == 0){
269 #No information on this particular GPL file 269 #No information on this particular GPL file
270 idLOCGPL <- genena %>% 270 idLOCGPL <- genena %>%
271 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>% 271 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>%
272 t(.) %>% 272 t(.) %>%
273 grep("^ID\\s*$",.) %>% 273 grep("^ID\\s*$",.) %>%
274 -1 274 -1
275 cbind(as.integer(gplnum),as.integer(idLOCGPL)) %>% 275 cbind(as.integer(gplnum),as.integer(idLOCGPL)) %>%
276 cat(file="GPL_ID_LOC.txt",sep = "\t", fill = TRUE, append = TRUE) 276 cat(file="GPL_ID_LOC.txt",sep = "\t", fill = TRUE, append = TRUE)
277 geneIDNam <- genena %>% 277 geneIDNam <- genena %>%
278 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>% 278 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>%
279 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 279 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
280 } 280 }
281 } else if(fileex == 0){ 281 } else if(fileex == 0){
282 #We must create a file that we can access for later use 282 #We must create a file that we can access for later use
283 idLOCGPL <- genena %>% 283 idLOCGPL <- genena %>%
284 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>% 284 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>%
285 t(.) %>% 285 t(.) %>%
286 grep("^ID\\s*$",.) %>% 286 grep("^ID\\s*$",.) %>%
287 -1 287 -1
288 Firstval <- cbind(as.integer(gplnum),as.integer(idLOCGPL)) 288 Firstval <- cbind(as.integer(gplnum),as.integer(idLOCGPL))
289 colnames(Firstval) <- c("GPL_FILE_NUM","LOC_ID") 289 colnames(Firstval) <- c("GPL_FILE_NUM","LOC_ID")
290 write.table(Firstval,file = "GPL_ID_LOC.txt", sep = "\t",row.names = FALSE, col.names = TRUE) 290 write.table(Firstval,file = "GPL_ID_LOC.txt", sep = "\t",row.names = FALSE, col.names = TRUE)
291 geneIDNam <- genena %>% 291 geneIDNam <- genena %>%
292 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>% 292 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>%
293 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 293 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
294 } 294 }
295 } else if(soft == FALSE){ 295 } else if(soft == FALSE){
296 geneIDNam <- genena %>% 296 geneIDNam <- genena %>%
297 read_delim(delim="\t",comment = "#")%>% 297 read_delim(delim="\t",comment = "#")%>%
298 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 298 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
299 } 299 }
300 300
301 ##Labeling the gene IDs without names 301 ##Labeling the gene IDs without names
302 geneIDNam <- NAFIXING(geneIDNam) 302 geneIDNam <- NAFIXING(geneIDNam)
303 303
304 ##remove the whitespace 304 ##remove the whitespace
305 geneIDNam <- t(rbind(str_trim(t(geneIDNam)[1,]),str_trim(t(geneIDNam)[2,]))) 305 geneIDNam <- t(rbind(str_trim(t(geneIDNam)[1,]),str_trim(t(geneIDNam)[2,])))
306 306
307 ##Here is the clean version 307 ##Here is the clean version
308 write.table(geneIDNam,file = paste0("Clean_GPL",gplnum,".txt"),sep = "\t",row.names = FALSE, col.names = FALSE) 308 write.table(geneIDNam,file = paste0("Clean_GPL",gplnum,".txt"),sep = "\t",row.names = FALSE, col.names = FALSE)
309 } 309 }
310 310
311 311
312 312
313 ##Changing the gene ID to gene name 313 ##Changing the gene ID to gene name
314 ALZDAT1 <- cgeneID(geneIDNam,alzdat) 314 ALZDAT1 <- cgeneID(geneIDNam,alzdat)
315 colnames(ALZDAT) = ALZDAT1[1,] 315 colnames(ALZDAT) = ALZDAT1[1,]
316 316
317 317
318 ##Adjusting the column names aka the gene names 318 ##Adjusting the column names aka the gene names
319 colnames(ALZDAT) <- gcnames(ALZDAT) 319 colnames(ALZDAT) <- gcnames(ALZDAT)
320 320
321 321
322 #Full RAW Data 322 #Full RAW Data
323 Fullalzdwr <- ALZDAT %>% 323 Fullalzdwr <- ALZDAT %>%
324 as.data.frame(.,stringsAsFactors = FALSE) %>% 324 as.data.frame(.,stringsAsFactors = FALSE) %>%
325 cbind(ALZWORDF,.) 325 cbind(ALZWORDF,.)
326 326
327 #Raw file is output 327 #Raw file is output
328 nfnaex <- strsplit(alz,"[\\]") %>% 328 nfnaex <- strsplit(alz,"[\\]") %>%
329 .[[1]] %>% 329 .[[1]] %>%
330 .[length(.)] %>% 330 .[length(.)] %>%
331 gsub("\\D","",.) %>% 331 gsub("\\D","",.) %>%
332 c("GSE",.,"aftexcel.txt") %>% 332 c("GSE",.,"aftexcel.txt") %>%
333 paste(collapse = "") 333 paste(collapse = "")
334 write.table(t(Fullalzdwr), file = nfnaex, sep = "\t") 334 write.table(t(Fullalzdwr), file = nfnaex, sep = "\t")
335 335
336 336
337 337
338 #Now for the discretization part 338 #Now for the discretization part
339 ##get the wordy part again 339 ##get the wordy part again
340 rawword <- t(ALZWORDF) 340 rawword <- t(ALZWORDF)
341 341
342 ##where is ID_REF located 342 ##where is ID_REF located
343 hereim <- grep("ID_REF",rownames(rawword)) 343 hereim <- grep("ID_REF",rownames(rawword))
344 344
345 ##Subject Names GSM... 345 ##Subject Names GSM...
346 subjnam <- rawword[hereim,] 346 subjnam <- rawword[hereim,]
347 347
348 ##Getting the names for the rows 348 ##Getting the names for the rows
349 namedarows <- rownames(rawword)[-hereim] %>% 349 namedarows <- rownames(rawword)[-hereim] %>%
350 as.data.frame(.,stringsAsFactors = FALSE) 350 as.data.frame(.,stringsAsFactors = FALSE)
351 RAWWORD <- rawword[-hereim,] %>% 351 RAWWORD <- rawword[-hereim,] %>%
352 as.data.frame(.,stringsAsFactors = FALSE) %>% 352 as.data.frame(.,stringsAsFactors = FALSE) %>%
353 bind_cols(namedarows,.) 353 bind_cols(namedarows,.)
354 z <- 1 354 z <- 1
355 naroww <- as.data.frame(rep(0,dim(RAWWORD)[1]),stringsAsFactors = FALSE) 355 naroww <- as.data.frame(rep(0,dim(RAWWORD)[1]),stringsAsFactors = FALSE)
356 for(z in 1:dim(RAWWORD)[1]){ 356 for(z in 1:dim(RAWWORD)[1]){
357 if(sum(is.na(RAWWORD[z,])) > 0){ 357 if(sum(is.na(RAWWORD[z,])) > 0){
358 naroww[z,1] <- as.integer(sum(is.na(RAWWORD[z,]))) 358 naroww[z,1] <- as.integer(sum(is.na(RAWWORD[z,])))
359 } 359 }
360 if(length(grep("NA",RAWWORD[z,])) > 0){ 360 if(length(grep("NA",RAWWORD[z,])) > 0){
361 naroww[z,1] <- as.integer(length(grep("NA",RAWWORD[z,]))) + naroww[z,1] 361 naroww[z,1] <- as.integer(length(grep("NA",RAWWORD[z,]))) + naroww[z,1]
362 } 362 }
363 z <- z + 1 363 z <- z + 1
364 } 364 }
365 365
366 colnames(naroww) <- "ROW_NAs" 366 colnames(naroww) <- "ROW_NAs"
367 RAWWORD <- bind_cols(RAWWORD,naroww) 367 RAWWORD <- bind_cols(RAWWORD,naroww)
368 368
369 369
370 roALZna <- t(ALZDAT) %>% 370 roALZna <- t(ALZDAT) %>%
371 rownames(.) %>% 371 rownames(.) %>%
372 as.data.frame(.,stringsAsFactors = FALSE) 372 as.data.frame(.,stringsAsFactors = FALSE)
373 colnames(roALZna) <- "ID_REF" 373 colnames(roALZna) <- "ID_REF"
374 374
375 RAWDAT <- t(ALZDAT) %>% 375 RAWDAT <- t(ALZDAT) %>%
376 as.data.frame(.,stringsAsFactors = FALSE) 376 as.data.frame(.,stringsAsFactors = FALSE)
377 colnames(RAWDAT) <- NULL 377 colnames(RAWDAT) <- NULL
378 rownames(RAWDAT) <- NULL 378 rownames(RAWDAT) <- NULL
379 379
380 RAWDAT2 <- RAWDAT %>% 380 RAWDAT2 <- RAWDAT %>%
381 cbind(roALZna,.) %>% 381 cbind(roALZna,.) %>%
382 dplyr::arrange(.,ID_REF) 382 dplyr::arrange(.,ID_REF)
383 383
384 ##Editing the file for R processing 384 ##Editing the file for R processing
385 RAWDATID <- RAWDAT2[,1] %>% 385 RAWDATID <- RAWDAT2[,1] %>%
386 as.matrix(.) 386 as.matrix(.)
387 387
388 RAWDATNUM <- RAWDAT2[,-1] %>% 388 RAWDATNUM <- RAWDAT2[,-1] %>%
389 mapply(.,FUN = as.numeric) %>% 389 mapply(.,FUN = as.numeric) %>%
390 t(.) 390 t(.)
391 391
392 ##Consolidating genes with the same name 392 ##Consolidating genes with the same name
393 ###create empty matrix of size equal to tabRDATID 393 ###create empty matrix of size equal to tabRDATID
394 tabRDATID <- table(RAWDATID) 394 tabRDATID <- table(RAWDATID)
395 NuRDATN <- matrix(0, nrow = dim(RAWDATNUM)[1], ncol = length(tabRDATID)) 395 NuRDATN <- matrix(0, nrow = dim(RAWDATNUM)[1], ncol = length(tabRDATID))
396 j <- 1 396 j <- 1
397 for(j in 1:length(tabRDATID)){ 397 for(j in 1:length(tabRDATID)){
398 ##Putting the ones without duplicates in their new homes 398 ##Putting the ones without duplicates in their new homes
399 if(tabRDATID[j] == 1){ 399 if(tabRDATID[j] == 1){
400 NuRDATN[,j] <- RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])] 400 NuRDATN[,j] <- RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])]
401 } else if(tabRDATID[j] > 1){ 401 } else if(tabRDATID[j] > 1){
402 ##Averaging duplicates and putting them in their new homes 402 ##Averaging duplicates and putting them in their new homes
403 NuRDATN[,j] <- rowMeans(RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])],na.rm = TRUE) 403 NuRDATN[,j] <- rowMeans(RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])],na.rm = TRUE)
404 } 404 }
405 j <- j + 1 405 j <- j + 1
406 } 406 }
407 407
408 ##Scaling the Data 408 ##Scaling the Data
409 scrawdat <- NuRDATN%>% 409 scrawdat <- NuRDATN%>%
410 scale() 410 scale()
411 attr(scrawdat,"scaled:center") <- NULL 411 attr(scrawdat,"scaled:center") <- NULL
412 attr(scrawdat,"scaled:scale") <- NULL 412 attr(scrawdat,"scaled:scale") <- NULL
413 colnames(scrawdat) <- rownames(tabRDATID) 413 colnames(scrawdat) <- rownames(tabRDATID)
414 414
415 ##Discretized the Data 415 ##Discretized the Data
416 dialzdat <- scrawdat %>% 416 dialzdat <- scrawdat %>%
417 dndat(.) %>% 417 dndat(.) %>%
418 t()%>% 418 t()%>%
419 as.data.frame(.,stringsAsFactors = FALSE) 419 as.data.frame(.,stringsAsFactors = FALSE)
420 colnames(dialzdat) <- rownames(RAWDATNUM) 420 colnames(dialzdat) <- rownames(RAWDATNUM)
421 421
422 ##setting "ID_REF" as a new variable 422 ##setting "ID_REF" as a new variable
423 geneNAM <- as.data.frame(as.matrix(rownames(dialzdat),ncol=1),stringsAsFactors = FALSE) 423 geneNAM <- as.data.frame(as.matrix(rownames(dialzdat),ncol=1),stringsAsFactors = FALSE)
424 colnames(geneNAM) <- "ID_REF" 424 colnames(geneNAM) <- "ID_REF"
425 rownames(dialzdat) <- NULL 425 rownames(dialzdat) <- NULL
426 dialzdat <-bind_cols(geneNAM,dialzdat) 426 dialzdat <-bind_cols(geneNAM,dialzdat)
427 427
428 ##NAs in a column 428 ##NAs in a column
429 x <- 2 429 x <- 2
430 nacol <- as.data.frame(t(rep(0,dim(dialzdat)[2])),stringsAsFactors = FALSE) 430 nacol <- as.data.frame(t(rep(0,dim(dialzdat)[2])),stringsAsFactors = FALSE)
431 nacol[1,1] = "COL_NAs" 431 nacol[1,1] = "COL_NAs"
432 for(x in 2:dim(dialzdat)[2]){ 432 for(x in 2:dim(dialzdat)[2]){
433 nacol[1,x] <- as.integer(sum(is.na(dialzdat[,x]))) 433 nacol[1,x] <- as.integer(sum(is.na(dialzdat[,x])))
434 x <- x + 1 434 x <- x + 1
435 } 435 }
436 colnames(nacol) <- colnames(dialzdat) 436 colnames(nacol) <- colnames(dialzdat)
437 dialzdat <- bind_rows(dialzdat,nacol) 437 dialzdat <- bind_rows(dialzdat,nacol)
438 438
439 ##NAs in a row 439 ##NAs in a row
440 y <- 1 440 y <- 1
441 narowd <- as.data.frame(rep(0,dim(dialzdat)[1]),stringsAsFactors = FALSE) 441 narowd <- as.data.frame(rep(0,dim(dialzdat)[1]),stringsAsFactors = FALSE)
442 for(y in 1:dim(dialzdat)[1]){ 442 for(y in 1:dim(dialzdat)[1]){
443 narowd[y,1] <- as.integer(sum(is.na(dialzdat[y,]))) 443 narowd[y,1] <- as.integer(sum(is.na(dialzdat[y,])))
444 y <- y + 1 444 y <- y + 1
445 } 445 }
446 colnames(narowd) <- "ROW_NAs" 446 colnames(narowd) <- "ROW_NAs"
447 dialzdat <- bind_cols(dialzdat,narowd) 447 dialzdat <- bind_cols(dialzdat,narowd)
448 colnames(dialzdat)[2:(dim(dialzdat)[2]-1)] <- subjnam 448 colnames(dialzdat)[2:(dim(dialzdat)[2]-1)] <- subjnam
449 colnames(RAWWORD) <- colnames(dialzdat) 449 colnames(RAWWORD) <- colnames(dialzdat)
450 ##converting to character so that the clinical can be brought together with discrete data 450 ##converting to character so that the clinical can be brought together with discrete data
451 k <- 2 451 k <- 2
452 for(k in 2:dim(dialzdat)[2]-1){ 452 for(k in 2:dim(dialzdat)[2]-1){
453 dialzdat[,k] <- as.character(dialzdat[,k]) 453 dialzdat[,k] <- as.character(dialzdat[,k])
454 k <- k + 1 454 k <- k + 1
455 } 455 }
456 #The End the full data 456 #The End the full data
457 Dscrtalzdw <- bind_rows(RAWWORD,dialzdat) 457 Dscrtalzdw <- bind_rows(RAWWORD,dialzdat)
458 458
459 #Produces Discrete file 459 #Produces Discrete file
460 nfnaex2 <- strsplit(alz,"[\\|/]") %>% 460 nfnaex2 <- strsplit(alz,"[\\|/]") %>%
461 .[[1]] %>% 461 .[[1]] %>%
462 .[length(.)] %>% 462 .[length(.)] %>%
463 gsub("\\D","",.) %>% 463 gsub("\\D","",.) %>%
464 c("GSE",.,"dscrt.txt") %>% 464 c("GSE",.,"dscrt.txt") %>%
465 paste(collapse = "") 465 paste(collapse = "")
466 write.table(Dscrtalzdw, file = nfnaex2, sep = "\t",col.names = TRUE,row.names = FALSE) 466 write.table(Dscrtalzdw, file = nfnaex2, sep = "\t",col.names = TRUE,row.names = FALSE)
467 n <- n +1 467 n <- n +1
468 } 468 }
469 } else if(numDAT == 2){ 469 } else if(numDAT == 2){
470 #CHOOSE A DATA FILE TO CLEAN OR SEVERAL DATA FILES TO CLEAN 470 #CHOOSE A DATA FILE TO CLEAN OR SEVERAL DATA FILES TO CLEAN
471 471
472 #All the files you want to analyze 472 #All the files you want to analyze
473 ANDIS <- select.list(choices = list.files()[GSEfileloc],multiple = TRUE, title = "Choose the file/files you want to analyze:") 473 ANDIS <- select.list(choices = list.files()[GSEfileloc],multiple = TRUE, title = "Choose the file/files you want to analyze:")
474 if(length(ANDIS) == 0){ 474 if(length(ANDIS) == 0){
475 #Spit out a warning 475 #Spit out a warning
476 warning("You did not select any files and so no cleaning will be performed") 476 warning("You did not select any files and so no cleaning will be performed")
477 } else{ 477 } else{
478 #indexing the data files 478 #indexing the data files
479 n <- 1 479 n <- 1
480 for(n in 1: length(ANDIS)){ 480 for(n in 1: length(ANDIS)){
481 alz <- ANDIS[n] 481 alz <- ANDIS[n]
482 482
483 #Working with the wordy part of the document 483 #Working with the wordy part of the document
484 alzword <- alz %>% 484 alzword <- alz %>%
485 read_delim(delim ="\t",comment = "!Series",col_names = FALSE)%>% 485 read_delim(delim ="\t",comment = "!Series",col_names = FALSE)%>%
486 filter(grepl("!Sample",X1))%>% 486 filter(grepl("!Sample",X1))%>%
487 filter(!grepl("!Sample_contact",X1)) 487 filter(!grepl("!Sample_contact",X1))
488 488
489 #Getting the GPL file 489 #Getting the GPL file
490 genena <- grep("_platform_id",alzword$X1) %>% 490 genena <- grep("_platform_id",alzword$X1) %>%
491 alzword$X2[.] %>% 491 alzword$X2[.] %>%
492 str_trim(.) %>% 492 str_trim(.) %>%
493 paste0("^",.,"\\D") %>% 493 paste0("^",.,"\\D") %>%
494 grep(.,list.files()) %>% 494 grep(.,list.files()) %>%
495 list.files()[.] 495 list.files()[.]
496 496
497 #Find out if it is a soft GPL file or not 497 #Find out if it is a soft GPL file or not
498 soft <- strsplit(genena,"[\\|/]") %>% 498 soft <- strsplit(genena,"[\\|/]") %>%
499 .[[1]] %>% 499 .[[1]] %>%
500 .[length(.)] %>% 500 .[length(.)] %>%
501 grepl("soft",.) 501 grepl("soft",.)
502 502
503 ##Changing row names and column names: 503 ##Changing row names and column names:
504 ALZWORD <- t(alzword) 504 ALZWORD <- t(alzword)
505 rownames(ALZWORD)=NULL 505 rownames(ALZWORD)=NULL
506 colnames(ALZWORD) <- colnames(ALZWORD,do.NULL=FALSE) 506 colnames(ALZWORD) <- colnames(ALZWORD,do.NULL=FALSE)
507 ALZWORD <- chngrownm(ALZWORD)[-1,] 507 ALZWORD <- chngrownm(ALZWORD)[-1,]
508 ALZWORD <- ALZWORD%>% 508 ALZWORD <- ALZWORD%>%
509 as.data.frame(.,stringsAsFactors = FALSE)%>% 509 as.data.frame(.,stringsAsFactors = FALSE)%>%
510 dplyr::select(-starts_with("col")) 510 dplyr::select(-starts_with("col"))
511 511
512 ##Reorganizing information within the columns and final clinical data 512 ##Reorganizing information within the columns and final clinical data
513 ALZWORDF <- cinfo(ALZWORD) 513 ALZWORDF <- cinfo(ALZWORD)
514 514
515 515
516 #Working with Actual Data part of file 516 #Working with Actual Data part of file
517 alzdat <- alz %>% 517 alzdat <- alz %>%
518 read_delim(delim="\t",col_names=TRUE,comment = "!",skip=1) 518 read_delim(delim="\t",col_names=TRUE,comment = "!",skip=1)
519 ALZDAT <- t(alzdat[,-1]) 519 ALZDAT <- t(alzdat[,-1])
520 rownames(ALZDAT)=NULL 520 rownames(ALZDAT)=NULL
521 521
522 ##Is there a clean version of the GPL file available? 522 ##Is there a clean version of the GPL file available?
523 gplnum <- strsplit(genena,"[\\|/]") %>% 523 gplnum <- strsplit(genena,"[\\|/]") %>%
524 .[[1]] %>% 524 .[[1]] %>%
525 .[length(.)] %>% 525 .[length(.)] %>%
526 gsub("\\D","",.) 526 gsub("\\D","",.)
527 clfileex <- sum(grepl(paste0("Clean_GPL",gplnum),list.files())) 527 clfileex <- sum(grepl(paste0("Clean_GPL",gplnum),list.files()))
528 if(clfileex >= 1){ 528 if(clfileex >= 1){
529 #use the clean version 529 #use the clean version
530 geneIDNam <- paste0("Clean_GPL",gplnum,".txt") %>% 530 geneIDNam <- paste0("Clean_GPL",gplnum,".txt") %>%
531 read_delim(delim="\t",col_names = c("ID","Symbol"), comment = "!") 531 read_delim(delim="\t",col_names = c("ID","Symbol"), comment = "!")
532 532
533 } else if(clfileex == 0){ 533 } else if(clfileex == 0){
534 ##Lets Create a clean version 534 ##Lets Create a clean version
535 535
536 ##Gene ID to Gene Name 536 ##Gene ID to Gene Name
537 if(soft == TRUE){ 537 if(soft == TRUE){
538 #Check to see if there is already a file containing information on soft files 538 #Check to see if there is already a file containing information on soft files
539 fileex <- sum(grepl("GPL_ID_LOC.txt",list.files())) 539 fileex <- sum(grepl("GPL_ID_LOC.txt",list.files()))
540 if(fileex == 1){ 540 if(fileex == 1){
541 #Check to see if this GPL soft file has been used before 541 #Check to see if this GPL soft file has been used before
542 IDF <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>% 542 IDF <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>%
543 .$GPL_FILE_NUM%>% 543 .$GPL_FILE_NUM%>%
544 grepl(gplnum,.) %>% 544 grepl(gplnum,.) %>%
545 sum() 545 sum()
546 if(IDF == 1){ 546 if(IDF == 1){
547 IDLOCAL <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>% 547 IDLOCAL <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>%
548 .$GPL_FILE_NUM%>% 548 .$GPL_FILE_NUM%>%
549 grep(gplnum,.) 549 grep(gplnum,.)
550 idlocgpl <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>% 550 idlocgpl <- read_delim("GPL_ID_LOC.txt",delim = "\t",col_names = TRUE) %>%
551 .$LOC_ID %>% 551 .$LOC_ID %>%
552 .[IDLOCAL] 552 .[IDLOCAL]
553 geneIDNam <- genena %>% 553 geneIDNam <- genena %>%
554 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idlocgpl) %>% 554 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idlocgpl) %>%
555 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 555 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
556 } else if(IDF == 0){ 556 } else if(IDF == 0){
557 #No information on this particular GPL file 557 #No information on this particular GPL file
558 idLOCGPL <- genena %>% 558 idLOCGPL <- genena %>%
559 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>% 559 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>%
560 t(.) %>% 560 t(.) %>%
561 grep("^ID\\s*$",.) %>% 561 grep("^ID\\s*$",.) %>%
562 -1 562 -1
563 cbind(as.integer(gplnum),as.integer(idLOCGPL)) %>% 563 cbind(as.integer(gplnum),as.integer(idLOCGPL)) %>%
564 cat(file="GPL_ID_LOC.txt",sep = "\t", fill = TRUE, append = TRUE) 564 cat(file="GPL_ID_LOC.txt",sep = "\t", fill = TRUE, append = TRUE)
565 geneIDNam <- genena %>% 565 geneIDNam <- genena %>%
566 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>% 566 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>%
567 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 567 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
568 } 568 }
569 } else if(fileex == 0){ 569 } else if(fileex == 0){
570 #We must create a file that we can access for later use 570 #We must create a file that we can access for later use
571 idLOCGPL <- genena %>% 571 idLOCGPL <- genena %>%
572 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>% 572 read_delim(delim="\t",col_names = FALSE, comment = "!", n_max = 1000) %>%
573 t(.) %>% 573 t(.) %>%
574 grep("^ID\\s*$",.) %>% 574 grep("^ID\\s*$",.) %>%
575 -1 575 -1
576 Firstval <- cbind(as.integer(gplnum),as.integer(idLOCGPL)) 576 Firstval <- cbind(as.integer(gplnum),as.integer(idLOCGPL))
577 colnames(Firstval) <- c("GPL_FILE_NUM","LOC_ID") 577 colnames(Firstval) <- c("GPL_FILE_NUM","LOC_ID")
578 write.table(Firstval,file = "GPL_ID_LOC.txt", sep = "\t",row.names = FALSE, col.names = TRUE) 578 write.table(Firstval,file = "GPL_ID_LOC.txt", sep = "\t",row.names = FALSE, col.names = TRUE)
579 geneIDNam <- genena %>% 579 geneIDNam <- genena %>%
580 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>% 580 read_delim(delim="\t",col_names = TRUE, comment = "!", skip = idLOCGPL) %>%
581 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 581 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
582 } 582 }
583 } else if(soft == FALSE){ 583 } else if(soft == FALSE){
584 geneIDNam <- genena %>% 584 geneIDNam <- genena %>%
585 read_delim(delim="\t",comment = "#")%>% 585 read_delim(delim="\t",comment = "#")%>%
586 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$",colnames(.))) 586 dplyr::select(.,ID,grep("Symbol|^ORF\\s*$|^gene_assignment\\s*$|^Gene symbol$",colnames(.)))
587 } 587 }
588 588
589 ##Labeling the gene IDs without names 589 ##Labeling the gene IDs without names
590 geneIDNam <- NAFIXING(geneIDNam) 590 geneIDNam <- NAFIXING(geneIDNam)
591 591
592 ##remove the whitespace 592 ##remove the whitespace
593 geneIDNam <- t(rbind(str_trim(t(geneIDNam)[1,]),str_trim(t(geneIDNam)[2,]))) 593 geneIDNam <- t(rbind(str_trim(t(geneIDNam)[1,]),str_trim(t(geneIDNam)[2,])))
594 594
595 ##Here is the clean version 595 ##Here is the clean version
596 write.table(geneIDNam,file = paste0("Clean_GPL",gplnum,".txt"),sep = "\t",row.names = FALSE, col.names = FALSE) 596 write.table(geneIDNam,file = paste0("Clean_GPL",gplnum,".txt"),sep = "\t",row.names = FALSE, col.names = FALSE)
597 } 597 }
598 598
599 599
600 600
601 ##Changing the gene ID to gene name 601 ##Changing the gene ID to gene name
602 ALZDAT1 <- cgeneID(geneIDNam,alzdat) 602 ALZDAT1 <- cgeneID(geneIDNam,alzdat)
603 colnames(ALZDAT) = ALZDAT1[1,] 603 colnames(ALZDAT) = ALZDAT1[1,]
604 604
605 605
606 ##Adjusting the column names aka the gene names 606 ##Adjusting the column names aka the gene names
607 colnames(ALZDAT) <- gcnames(ALZDAT) 607 colnames(ALZDAT) <- gcnames(ALZDAT)
608 608
609 609
610 #Full RAW Data 610 #Full RAW Data
611 Fullalzdwr <- ALZDAT %>% 611 Fullalzdwr <- ALZDAT %>%
612 as.data.frame(.,stringsAsFactors = FALSE) %>% 612 as.data.frame(.,stringsAsFactors = FALSE) %>%
613 cbind(ALZWORDF,.) 613 cbind(ALZWORDF,.)
614 614
615 #Raw file is output 615 #Raw file is output
616 nfnaex <- strsplit(alz,"[\\]") %>% 616 nfnaex <- strsplit(alz,"[\\]") %>%
617 .[[1]] %>% 617 .[[1]] %>%
618 .[length(.)] %>% 618 .[length(.)] %>%
619 gsub("\\D","",.) %>% 619 gsub("\\D","",.) %>%
620 c("GSE",.,"aftexcel.txt") %>% 620 c("GSE",.,"aftexcel.txt") %>%
621 paste(collapse = "") 621 paste(collapse = "")
622 write.table(t(Fullalzdwr), file = nfnaex, sep = "\t") 622 write.table(t(Fullalzdwr), file = nfnaex, sep = "\t")
623 623
624 624
625 625
626 #Now for the discretization part 626 #Now for the discretization part
627 ##get the wordy part again 627 ##get the wordy part again
628 rawword <- t(ALZWORDF) 628 rawword <- t(ALZWORDF)
629 629
630 ##where is ID_REF located 630 ##where is ID_REF located
631 hereim <- grep("ID_REF",rownames(rawword)) 631 hereim <- grep("ID_REF",rownames(rawword))
632 632
633 ##Subject Names GSM... 633 ##Subject Names GSM...
634 subjnam <- rawword[hereim,] 634 subjnam <- rawword[hereim,]
635 635
636 ##Getting the names for the rows 636 ##Getting the names for the rows
637 namedarows <- rownames(rawword)[-hereim] %>% 637 namedarows <- rownames(rawword)[-hereim] %>%
638 as.data.frame(.,stringsAsFactors = FALSE) 638 as.data.frame(.,stringsAsFactors = FALSE)
639 RAWWORD <- rawword[-hereim,] %>% 639 RAWWORD <- rawword[-hereim,] %>%
640 as.data.frame(.,stringsAsFactors = FALSE) %>% 640 as.data.frame(.,stringsAsFactors = FALSE) %>%
641 bind_cols(namedarows,.) 641 bind_cols(namedarows,.)
642 z <- 1 642 z <- 1
643 naroww <- as.data.frame(rep(0,dim(RAWWORD)[1]),stringsAsFactors = FALSE) 643 naroww <- as.data.frame(rep(0,dim(RAWWORD)[1]),stringsAsFactors = FALSE)
644 for(z in 1:dim(RAWWORD)[1]){ 644 for(z in 1:dim(RAWWORD)[1]){
645 if(sum(is.na(RAWWORD[z,])) > 0){ 645 if(sum(is.na(RAWWORD[z,])) > 0){
646 naroww[z,1] <- as.integer(sum(is.na(RAWWORD[z,]))) 646 naroww[z,1] <- as.integer(sum(is.na(RAWWORD[z,])))
647 } 647 }
648 if(length(grep("NA",RAWWORD[z,])) > 0){ 648 if(length(grep("NA",RAWWORD[z,])) > 0){
649 naroww[z,1] <- as.integer(length(grep("NA",RAWWORD[z,]))) + naroww[z,1] 649 naroww[z,1] <- as.integer(length(grep("NA",RAWWORD[z,]))) + naroww[z,1]
650 } 650 }
651 z <- z + 1 651 z <- z + 1
652 } 652 }
653 653
654 colnames(naroww) <- "ROW_NAs" 654 colnames(naroww) <- "ROW_NAs"
655 RAWWORD <- bind_cols(RAWWORD,naroww) 655 RAWWORD <- bind_cols(RAWWORD,naroww)
656 656
657 657
658 roALZna <- t(ALZDAT) %>% 658 roALZna <- t(ALZDAT) %>%
659 rownames(.) %>% 659 rownames(.) %>%
660 as.data.frame(.,stringsAsFactors = FALSE) 660 as.data.frame(.,stringsAsFactors = FALSE)
661 colnames(roALZna) <- "ID_REF" 661 colnames(roALZna) <- "ID_REF"
662 662
663 RAWDAT <- t(ALZDAT) %>% 663 RAWDAT <- t(ALZDAT) %>%
664 as.data.frame(.,stringsAsFactors = FALSE) 664 as.data.frame(.,stringsAsFactors = FALSE)
665 colnames(RAWDAT) <- NULL 665 colnames(RAWDAT) <- NULL
666 rownames(RAWDAT) <- NULL 666 rownames(RAWDAT) <- NULL
667 667
668 RAWDAT2 <- RAWDAT %>% 668 RAWDAT2 <- RAWDAT %>%
669 cbind(roALZna,.) %>% 669 cbind(roALZna,.) %>%
670 dplyr::arrange(.,ID_REF) 670 dplyr::arrange(.,ID_REF)
671 671
672 ##Editing the file for R processing 672 ##Editing the file for R processing
673 RAWDATID <- RAWDAT2[,1] %>% 673 RAWDATID <- RAWDAT2[,1] %>%
674 as.matrix(.) 674 as.matrix(.)
675 675
676 RAWDATNUM <- RAWDAT2[,-1] %>% 676 RAWDATNUM <- RAWDAT2[,-1] %>%
677 mapply(.,FUN = as.numeric) %>% 677 mapply(.,FUN = as.numeric) %>%
678 t(.) 678 t(.)
679 679
680 ##Consolidating genes with the same name 680 ##Consolidating genes with the same name
681 ###create empty matrix of size equal to tabRDATID 681 ###create empty matrix of size equal to tabRDATID
682 tabRDATID <- table(RAWDATID) 682 tabRDATID <- table(RAWDATID)
683 NuRDATN <- matrix(0, nrow = dim(RAWDATNUM)[1], ncol = length(tabRDATID)) 683 NuRDATN <- matrix(0, nrow = dim(RAWDATNUM)[1], ncol = length(tabRDATID))
684 j <- 1 684 j <- 1
685 for(j in 1:length(tabRDATID)){ 685 for(j in 1:length(tabRDATID)){
686 ##Putting the ones without duplicates in their new homes 686 ##Putting the ones without duplicates in their new homes
687 if(tabRDATID[j] == 1){ 687 if(tabRDATID[j] == 1){
688 NuRDATN[,j] <- RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])] 688 NuRDATN[,j] <- RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])]
689 } else if(tabRDATID[j] > 1){ 689 } else if(tabRDATID[j] > 1){
690 ##Averaging duplicates and putting them in their new homes 690 ##Averaging duplicates and putting them in their new homes
691 NuRDATN[,j] <- rowMeans(RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])],na.rm = TRUE) 691 NuRDATN[,j] <- rowMeans(RAWDATNUM[,which(RAWDATID==rownames(tabRDATID)[j])],na.rm = TRUE)
692 } 692 }
693 j <- j + 1 693 j <- j + 1
694 } 694 }
695 695
696 ##Scaling the Data 696 ##Scaling the Data
697 scrawdat <- NuRDATN%>% 697 scrawdat <- NuRDATN%>%
698 scale() 698 scale()
699 attr(scrawdat,"scaled:center") <- NULL 699 attr(scrawdat,"scaled:center") <- NULL
700 attr(scrawdat,"scaled:scale") <- NULL 700 attr(scrawdat,"scaled:scale") <- NULL
701 colnames(scrawdat) <- rownames(tabRDATID) 701 colnames(scrawdat) <- rownames(tabRDATID)
702 702
703 ##Discretized the Data 703 ##Discretized the Data
704 dialzdat <- scrawdat %>% 704 dialzdat <- scrawdat %>%
705 dndat(.) %>% 705 dndat(.) %>%
706 t()%>% 706 t()%>%
707 as.data.frame(.,stringsAsFactors = FALSE) 707 as.data.frame(.,stringsAsFactors = FALSE)
708 colnames(dialzdat) <- rownames(RAWDATNUM) 708 colnames(dialzdat) <- rownames(RAWDATNUM)
709 709
710 ##setting "ID_REF" as a new variable 710 ##setting "ID_REF" as a new variable
711 geneNAM <- as.data.frame(as.matrix(rownames(dialzdat),ncol=1),stringsAsFactors = FALSE) 711 geneNAM <- as.data.frame(as.matrix(rownames(dialzdat),ncol=1),stringsAsFactors = FALSE)
712 colnames(geneNAM) <- "ID_REF" 712 colnames(geneNAM) <- "ID_REF"
713 rownames(dialzdat) <- NULL 713 rownames(dialzdat) <- NULL
714 dialzdat <-bind_cols(geneNAM,dialzdat) 714 dialzdat <-bind_cols(geneNAM,dialzdat)
715 715
716 ##NAs in a column 716 ##NAs in a column
717 x <- 2 717 x <- 2
718 nacol <- as.data.frame(t(rep(0,dim(dialzdat)[2])),stringsAsFactors = FALSE) 718 nacol <- as.data.frame(t(rep(0,dim(dialzdat)[2])),stringsAsFactors = FALSE)
719 nacol[1,1] = "COL_NAs" 719 nacol[1,1] = "COL_NAs"
720 for(x in 2:dim(dialzdat)[2]){ 720 for(x in 2:dim(dialzdat)[2]){
721 nacol[1,x] <- as.integer(sum(is.na(dialzdat[,x]))) 721 nacol[1,x] <- as.integer(sum(is.na(dialzdat[,x])))
722 x <- x + 1 722 x <- x + 1
723 } 723 }
724 colnames(nacol) <- colnames(dialzdat) 724 colnames(nacol) <- colnames(dialzdat)
725 dialzdat <- bind_rows(dialzdat,nacol) 725 dialzdat <- bind_rows(dialzdat,nacol)
726 726
727 ##NAs in a row 727 ##NAs in a row
728 y <- 1 728 y <- 1
729 narowd <- as.data.frame(rep(0,dim(dialzdat)[1]),stringsAsFactors = FALSE) 729 narowd <- as.data.frame(rep(0,dim(dialzdat)[1]),stringsAsFactors = FALSE)
730 for(y in 1:dim(dialzdat)[1]){ 730 for(y in 1:dim(dialzdat)[1]){
731 narowd[y,1] <- as.integer(sum(is.na(dialzdat[y,]))) 731 narowd[y,1] <- as.integer(sum(is.na(dialzdat[y,])))
732 y <- y + 1 732 y <- y + 1
733 } 733 }
734 colnames(narowd) <- "ROW_NAs" 734 colnames(narowd) <- "ROW_NAs"
735 dialzdat <- bind_cols(dialzdat,narowd) 735 dialzdat <- bind_cols(dialzdat,narowd)
736 colnames(dialzdat)[2:(dim(dialzdat)[2]-1)] <- subjnam 736 colnames(dialzdat)[2:(dim(dialzdat)[2]-1)] <- subjnam
737 colnames(RAWWORD) <- colnames(dialzdat) 737 colnames(RAWWORD) <- colnames(dialzdat)
738 ##converting to character so that the clinical can be brought together with discrete data 738 ##converting to character so that the clinical can be brought together with discrete data
739 k <- 2 739 k <- 2
740 for(k in 2:dim(dialzdat)[2]-1){ 740 for(k in 2:dim(dialzdat)[2]-1){
741 dialzdat[,k] <- as.character(dialzdat[,k]) 741 dialzdat[,k] <- as.character(dialzdat[,k])
742 k <- k + 1 742 k <- k + 1
743 } 743 }
744 #The End the full data 744 #The End the full data
745 Dscrtalzdw <- bind_rows(RAWWORD,dialzdat) 745 Dscrtalzdw <- bind_rows(RAWWORD,dialzdat)
746 746
747 #Produces Discrete file 747 #Produces Discrete file
748 nfnaex2 <- strsplit(alz,"[\\|/]") %>% 748 nfnaex2 <- strsplit(alz,"[\\|/]") %>%
749 .[[1]] %>% 749 .[[1]] %>%
750 .[length(.)] %>% 750 .[length(.)] %>%
751 gsub("\\D","",.) %>% 751 gsub("\\D","",.) %>%
752 c("GSE",.,"dscrt.txt") %>% 752 c("GSE",.,"dscrt.txt") %>%
753 paste(collapse = "") 753 paste(collapse = "")
754 write.table(Dscrtalzdw, file = nfnaex2, sep = "\t",col.names = TRUE,row.names = FALSE) 754 write.table(Dscrtalzdw, file = nfnaex2, sep = "\t",col.names = TRUE,row.names = FALSE)
755 755
756 756
757 n <- n + 1 757 n <- n + 1
758 } 758 }
759 } 759 }
760 } 760 }
761 } 761 }
762 #The Rest of this code will be used every time you want to change a data set 762 #The Rest of this code will be used every time you want to change a data set
763 THEFT() 763 THEFT()