Add files via upload

00_merge_area-size.Rmd

@@ -0,0 +1,230 @@
+---
+title: "R Notebook"
+output: html_notebook
+---
+
+```{r}
+wd <- dirname(getwd())
+
+#Set working directory and folder structure
+data_folder <- file.path(wd,'clinical_data')
+
+#set plot folder for results
+plot_folder <- file.path(wd,"results","merged")
+
+set.seed(101100)
+
+```
+
+
+```{r}
+wd <-dirname(getwd())
+
+t86a <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_86A.csv")))
+t86b <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_86B.csv")))
+t86c <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_86C.csv")))
+
+t87a <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_87A.csv")))
+t87b <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_87B.csv")))
+t87c <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_87C.csv")))
+
+t88a <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_88A.csv")))
+t88b <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_88B.csv")))
+t88c <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_88C.csv")))
+
+t175a <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_175A.csv")))
+t175b <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_175B.csv")))
+t175c <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_175C.csv")))
+
+t176a <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_176A.csv")))
+t176b <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_176B.csv")))
+t176c <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_176C.csv")))
+
+t178a <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_178A.csv")))
+t178b <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_178B.csv")))
+t178c <- read.csv(file=file.path(wd,"clinical_data", paste("Resized_AreaMeasurements_Image_178C.csv")))
+
+
+t86a$TMA <- "86A"
+t86b$TMA <- "86B"
+t86c$TMA <- "86C"
+
+t87a$TMA <- "87A"
+t87b$TMA <- "87B"
+t87c$TMA <- "87C"
+
+t88a$TMA <- "88A"
+t88b$TMA <- "88B"
+t88c$TMA <- "88C"
+
+t175a$TMA <- "175A"
+t175b$TMA <- "175B"
+t175c$TMA <- "175C"
+
+t176a$TMA <- "176A"
+t176b$TMA <- "176B"
+t176c$TMA <- "176C"
+
+t178a$TMA <- "178A"
+t178b$TMA <- "178B"
+t178c$TMA <- "178C"
+
+t.area <-rbind(t86a,t86b,t86c,t87a,t87b,t87c,t88a,t88b,t88c,t175a,t175b,t175c,t176a,t176b,t176c,t178a,t178b,t178c)
+t.area$ImageID <- t.area$Metadata_acid
+t.area <- t.area %>% select(c(AreaOccupied_AreaOccupied_StromaBinary,AreaOccupied_AreaOccupied_TumourBinary,AreaOccupied_AreaOccupied_TumourStromaBinary,TMA, ImageID)) #ImageNumber
+
+t.area <-t.area %>% unite("TMA_ImageID", c(TMA, ImageID), remove=T)
+colnames(t.area) <- c("Area_px_Stroma","Area_px_Tumour","Area_px_Core", "TMA_ImageID")
+
+range(t.area$Area_px_Core)/1000000
+hist(t.area$Area_px_Core/1000000)
+summary(t.area$Area_px_Core/1000000)
+t.area.mm <-
+  t.area %>% 
+  dplyr::mutate_at(vars(Area_px_Stroma:Area_px_Core),
+            .funs = funs(. /1000000))
+
+t.area.mm[t.area.mm$Area_px_Core>1,]
+t.area.mm$ImageID <-NULL
+colnames(t.area.mm) <- c("Area_mm_Stroma","Area_mm_Tumour","Area_mm_Core", "TMA_ImageID")
+area <-left_join(t.area, t.area.mm, by="TMA_ImageID")
+
+write.csv(area, file=file.path(data_folder, "area.csv"))
+```
+
+
+```{r}
+colData(roi.sce) <-as.data.frame(colData(roi.sce)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+colData(pat.sce) <-as.data.frame(colData(pat.sce)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+colData(roi.sce) <-as.data.frame(colData(roi.sce)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+colData(roi.pat.sce) <-as.data.frame(colData(roi.pat.sce)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+colData(immune.u1) <-as.data.frame(colData(immune.u1)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+colData(tumour.final) <-as.data.frame(colData(tumour.final)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+
+colData(vessel.sce) <-as.data.frame(colData(vessel.sce)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+colData(vessel.vessel_pat.sce) <-as.data.frame(colData(vessel.vessel_pat.sce)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+
+colData(tumour.final) <-as.data.frame(colData(tumour.final)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+#immune
+immune.o <-immune.final
+immune.final <-immune.o
+colData(immune.final) <-as.data.frame(colData(immune.final)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+dat.sce <- as_tibble(colData(immune.final))
+
+dat.sce<-dat.sce %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE)
+
+dat.sce <-DataFrame(dat.sce)
+dat.sce<-merge(dat.sce, area, by="TMA_ImageID")
+rownames(dat.sce) <-paste(dat.sce$TmaID, dat.sce$TmaBlock, dat.sce$ImageID, dat.sce$CellNumber, sep='_')
+colData(immune.final) <- dat.sce
+colnames(immune.final) <-rownames(dat.sce)
+colData(immune.final)
+colnames(colData(immune.final))
+  
+#tcells
+tcells.o <-tcells.final
+tcells.final <-tcells.o
+colData(tcells.final) <-as.data.frame(colData(tcells.final)) %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE) %>% left_join(area, by="TMA_ImageID") %>% DataFrame()
+
+
+dat.sce <- as_tibble(colData(tcells.final))
+dat.sce$Area_mm_Core <-NULL
+dat.sce$Area_mm_Stroma <-NULL
+dat.sce$Area_mm_Tumour <-NULL
+dat.sce$Area_px_Core <-NULL
+dat.sce$Area_px_Stroma <-NULL
+dat.sce$Area_px_Tumour <-NULL
+dat.sce$TMA_Image <-NULL
+dat.sce<-dat.sce %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE)
+
+dat.sce <-DataFrame(dat.sce)
+dat.sce<-merge(dat.sce, area, by="TMA_ImageID")
+rownames(dat.sce) <-paste(dat.sce$TmaID, dat.sce$TmaBlock, dat.sce$ImageID, dat.sce$CellNumber, sep='_')
+colData(tcells.final) <- dat.sce
+colnames(tcells.final) <-rownames(dat.sce)
+colData(tcells.final)
+colnames(colData(tcells.final))
+
+#tumour
+dat.sce <- as_tibble(colData(tumour))
+
+dat.sce<-dat.sce %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE)
+
+dat.sce <-DataFrame(dat.sce)
+dat.sce<-merge(dat.sce, area, by="TMA_ImageID")
+rownames(dat.sce) <-paste(dat.sce$TmaID, dat.sce$TmaBlock, dat.sce$ImageID, dat.sce$CellNumber, sep='_')
+colData(tumour) <- dat.sce
+colnames(tumour) <-rownames(dat.sce)
+colData(tumour)
+colnames(colData(tumour))
+  
+#fibro
+dat.sce <- as_tibble(colData(fibro.final.new))
+
+dat.sce<-dat.sce %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE)
+
+dat.sce <-DataFrame(dat.sce)
+dat.sce<-merge(dat.sce, area, by="TMA_ImageID")
+rownames(dat.sce) <-paste(dat.sce$TmaID, dat.sce$TmaBlock, dat.sce$ImageID, dat.sce$CellNumber, sep='_')
+colData(fibro.final.new) <- dat.sce
+colnames(fibro.final.new) <-rownames(dat.sce)
+colData(fibro.final.new)
+colnames(colData(fibro.final.new))
+  
+#vessel
+dat.sce <- as_tibble(colData(vessel))
+
+dat.sce<-dat.sce %>% 
+  unite(TMA_ImageID, c(TMA, ImageID), sep = "_", remove = FALSE)
+
+dat.sce <-DataFrame(dat.sce)
+dat.sce<-merge(dat.sce, area, by="TMA_ImageID")
+rownames(dat.sce) <-paste(dat.sce$TmaID, dat.sce$TmaBlock, dat.sce$ImageID, dat.sce$CellNumber, sep='_')
+colData(vessel) <- dat.sce
+colnames(vessel) <-rownames(dat.sce)
+colData(vessel)
+colnames(colData(vessel))
+length(unique(tumour[,tumour$TMA=="88_A"]$ImageID))  
+order(unique(tumour[,tumour$TMA=="88_A"]$ImageID))
+fibro$Area_mm
+```
+
+```{r}
+wd <- dirname(getwd())
+
+#Set working directory and folder structure
+data_folder <- file.path(wd,'TMA_all',"SCE")
+
+
+set.seed(101100)
+
+saveRDS(tumour,file=file.path(data_folder,paste("tumour_sce_merge_minus_ctrl_area.rds")))
+saveRDS(immune,file=file.path(data_folder,paste("immune_sce_merge_minus_ctrl_area.rds")))
+saveRDS(tcells,file=file.path(data_folder,paste("tcells_sce_merge_minus_ctrl_area.rds")))
+saveRDS(fibro,file=file.path(data_folder,paste("fibro_sce_merge_minus_ctrl_area.rds")))
+saveRDS(vessel,file=file.path(data_folder,paste("vessel_sce_merge_minus_ctrl_area.rds")))
+```

00_prepare_clinicaldata.Rmd

@@ -0,0 +1,139 @@
+---
+title: "Prepare clinical data and position data"
+output: html_notebook
+---
+```{r install needed packages}
+#install.packages("zoo")              
+library("zoo")    
+```
+
+```{r Set file structure}
+wd <- (getwd())
+data_folder <- file.path(dirname(wd),'clinical_data')
+```
+
+
+```{r Read in clinical data}
+clinical.data <-read.csv(file=file.path(data_folder, paste("TMA_86_87_88_175_176_178_USZ.csv")))
+```
+
+```{r Read in position files}
+p86 <- read.csv(file=file.path(data_folder, paste("zTMA_86.csv")),header=T, na.strings=c("","NA"))
+p87 <- read.csv(file=file.path(data_folder, paste("zTMA_87.csv")),header=T, na.strings=c("","NA"))
+p88 <- read.csv(file=file.path(data_folder, paste("zTMA_88.csv")),header=T, na.strings=c("","NA"))
+
+p175 <-read.csv(file=file.path(data_folder, paste("zTMA_175.csv")),header=T, na.strings=c("","NA"))
+p176 <-read.csv(file=file.path(data_folder, paste("zTMA_176.csv")),header=T, na.strings=c("","NA"))
+p178 <- read.csv(file=file.path(data_folder, paste("zTMA_178.csv")),header=T, na.strings=c("","NA"))
+```
+
+```{r Add missing patient numbers and ctrls}
+p86$Patient_Nr <-na.locf(p86$Patient_Nr)
+p87$Patient_Nr <-na.locf(p87$Patient_Nr)
+p88$Patient_Nr <-na.locf(p88$Patient_Nr)
+
+p175$Patient_Nr <-na.locf(p175$Patient_Nr)
+p176$Patient_Nr <-na.locf(p176$Patient_Nr)
+p178$Patient_Nr <-na.locf(p178$Patient_Nr)
+```
+
+```{r Add TMA column to position data}
+p86["TMA"] <-"86"
+p87["TMA"] <-"87"
+p88["TMA"] <-"88"
+
+p175["TMA"] <-"175"
+p176["TMA"] <-"176"
+p178["TMA"] <-"178"
+```
+
+```{r merge clinical data and position data}
+clinical.data_pos <-clinical.data
+
+#change Nr. to Patient_Nr
+names(clinical.data_pos)[names(clinical.data_pos) == 'Nr.'] <- 'Patient_Nr'
+
+#combine all position files rowwise
+p_combined <- rbind(p86,p87,p88,p175,p176,p178)
+
+clinical.data_pos <-merge(p_combined, clinical.data_pos, by.x=c("Patient_Nr", "TMA"), by.y=c("Patient_Nr","TMA")) 
+```
+
+```{r save combined clincal and position data as csv file}
+#write.csv(clinical.data_pos, file=file.path(data_folder,paste("combined_clinical_position_data.csv")))
+
+clinical.data <- read.csv( file=file.path(data_folder,paste("combined_clinical_position_data.csv")))
+
+table(clinical.data$DX.name)
+clinical.data$DX.name[clinical.data$DX.name == "Adeno-Ca"] <- "Adenocarcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "Adeno-Ca "] <- "Adenocarcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "PE-Ca"] <- "Squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "adenosquam. Ca"] <- "Adeno squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "ASQ"] <- "Adeno squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "Adeno squamous cell carcinoma"] <- "Adeno squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "Adenosquamöses CA"] <- "Adeno squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "Adeno/squamous carcinoma"] <- "Adeno squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "LC"] <- "Large cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "SCC"] <- "Squamous cell carcinoma"
+clinical.data$DX.name[clinical.data$DX.name == "LCNEC"] <- "Large cell neuroendocrine carcinoma"
+
+unique(clinical.data$DX.name)
+#add factor levels
+levels(clinical.data$Gender) <- list(male=1, female=2)
+levels(clinical.data$Grade) <- list("missing"=0,"Grade 1"=1,  "Grade 2"=2,"Grade 3"=3)
+levels(clinical.data$Vessel) <- list("negative"=0, "positive"=1)
+levels(clinical.data$Pleura) <- list("negative"=0, "positive"=1)
+levels(clinical.data$R) <- list("R0"=0, "R1"=1)
+levels(clinical.data$Relapse) <- list("no"=0, "yes"=1)
+levels(clinical.data$Ev.O) <- list("alive"=0, "dead"=1)
+levels(clinical.data$T.new )<- list("1a"=1, "1b"=2,"2a"=3, "2b"=4, "3"=5, "4"=6)
+levels(clinical.data$M.new) <- list("no"=0, "1a"=1, "1b"=2)
+levels(clinical.data$N) <- list("N0"=0, "N1"=1,"N2"=2, "N3"=3)
+
+levels(clinical.data$Stage) <- list("1a"=1, "1b"=2,"2a"=3, "2b"=4, "3a"=5, "3b"=6,"4"=7)
+levels(clinical.data$Chemo) <- list("Neoadjuvant no"=0, "Neoadjuvant yes"=1)
+levels(clinical.data$Radio) <- list("Neoadjuvant no"=0, "Neoadjuvant yes"=1)
+levels(clinical.data$Chemo3) <- list("Adjuvant no"=0, "Adjuvant yes"=1)
+levels(clinical.data$Radio4) <- list("Adjuvant no"=0, "Adjuvant loc"=1, "Other"=2, "Both"=3)
+levels(clinical.data$Chemo3) <- list("Post relapse no"=0, "Post relapse yes"=1)
+levels(clinical.data$Radio6) <- list("Post relapse no"=0, "Post relapse loc"=1, "Post relapse other"=2, "Post relapse both"=3)
+levels(clinical.data$Smok) <- list("no"=0, "currently"=1, "former"=2, "unknown"=3)
+
+cols <- c("TMA","Gender","Typ","Grade","Vessel","Pleura","T.new","N","M.new","Stage","R","Chemo","Radio","Chemo3","Radio4","Relapse","Chemo5","Radio6","DFS","Ev.O" ,"Smok"   )
+clinical.data[cols] <- lapply(clinical.data[cols], as.factor)
+
+head(clinical.data)
+
+clinical.data$ROI <- paste(clinical.data$Grid, clinical.data$x.y.localisation, sep="")
+head(clinical.data$ROI)
+clinical.data$RoiID <-paste(clinical.data$TMA, clinical.data$ROI, sep="_")
+clinical.data$Patient_ID <- paste(clinical.data$TMA, clinical.data$Patient_Nr,sep="_")
+
+
+clinical.data$LN.Met <- ifelse(clinical.data$N ==0, "No LN Metastases", "LN Metastases")
+clinical.data$Dist.Met <- ifelse(clinical.data$M.new ==0, "No Dist. Metastases", "Dist. Metastases")
+clinical.data$NeoAdj <- ifelse(clinical.data$Radio==1 |clinical.data$Chemo==1, "NeoAdjuvantTherapy", "NoNeoAdjuvantTherapy")
+clinical.data$X <-NULL
+clinical.data$TmaBlock <- clinical.data$Grid
+clinical.data$Grid <-NULL
+clinical.data$x.localisation <-NULL
+clinical.data$y.localisation <-NULL
+
+wd <-"/mnt"
+
+data_folder <-(file.path(wd,"lena_processed2","NSCLC_NEW","clinical_data"))
+
+write.csv(clinical.data, file=file.path(data_folder,"clinical_data.csv"))
+```
+
+
+```{r save position files}
+write.csv(p86, file=file.path(data_folder, paste("zTMA_86.csv")))
+write.csv(p87, file=file.path(data_folder, paste("zTMA_87.csv")))
+write.csv(p88, file=file.path(data_folder, paste("zTMA_88.csv")))
+
+write.csv(p175, file=file.path(data_folder, paste("zTMA_175.csv")))
+write.csv(p176, file=file.path(data_folder, paste("zTMA_176.csv")))
+write.csv(p178, file=file.path(data_folder, paste("zTMA_178.csv")))
+```
+