Add/update singler scmatch code

r.scripts/SingleR.R

+load("./genesets/scDWS.250.rda")
+load("./genesets/cibersort.rda")
+
+singler = CreateSinglerObject(GetAssayData(sc10x,assay="RNA"), annot = NULL, "PdPbPc_deep", min.genes = 0,
+                              technology = "10X", species = "Human", citation = "",
+                              ref.list = list(scDWS.250), normalize.gene.length = F, variable.genes = "de",
+                              fine.tune = T, do.signatures = T, clusters = NULL, do.main.types = T, 
+                              reduce.file.size = T, numCores = SingleR.numCores)
+
+cells.epi <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single.main$labels[,1]=="Epi",])
+cells.fmst <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single.main$labels[,1]=="FMSt",])
+cells.endo <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single.main$labels[,1]=="Endo",])
+cells.st <- c(cells.fmst,cells.endo)
+cells.leu <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single.main$labels[,1]=="Leu",])
+
+cells.be <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="BE",])
+cells.le <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="LE",])
+cells.hillock <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="Hillock",])
+cells.club <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="Club",])
+cells.ne <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="NE",])
+cells.fib <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="Fib",])
+cells.sm <-names(singler$singler[[1]]$SingleR.single.main$labels[singler$singler[[1]]$SingleR.single$labels[,1]=="SM",])
+
+Idents(sc10x,cells=cells.epi) <- "Epi"
+Idents(sc10x,cells=cells.fmst) <- "FMSt"
+Idents(sc10x,cells=cells.endo) <- "Endo"
+Idents(sc10x,cells=cells.leu) <- "Leu"
+sc10x$lin <- Idents(sc10x)
+
+sc10x.epi <- subset(sc10x,idents="Epi")
+sc10x.fmst <- subset(sc10x,idents="FMSt")
+sc10x.st <- subset(sc10x,idents=c("FMSt","Endo"))
+sc10x.leu <- subset(sc10x,idents="Leu")
+
+Idents(sc10x.epi,cells=cells.be) <- "BE"
+Idents(sc10x.epi,cells=cells.le) <- "LE"
+Idents(sc10x.epi,cells=cells.hillock) <- "Hillock"
+Idents(sc10x.epi,cells=cells.club) <- "Club"
+Idents(sc10x.epi,cells=cells.ne) <- "NE"
+sc10x.epi$epi.pops <- Idents(sc10x.epi)
+Idents(sc10x.fmst,cells=cells.fib) <- "Fib"
+Idents(sc10x.fmst,cells=cells.sm) <- "SM"
+sc10x.fmst$fmst.pops <- Idents(sc10x.fmst)
+Idents(sc10x.st,cells=cells.fib) <- "Fib"
+Idents(sc10x.st,cells=cells.sm) <- "SM"
+Idents(sc10x.st,cells=cells.endo) <- "Endo"
+sc10x.st$st.pops <- Idents(sc10x.st)
+
+results <- scPC(sc10x.epi,pc=100,hpc=0.9,file="epi",print="2")
+sc10x.epi <- results[[1]]
+pc.use.epi <- results[[2]]
+rm(results)
+sc10x.epi <- scCluster(sc10x.epi,res=res,red="pca",dim=pc.use.epi,print="2",folder="epi")
+
+results <- scPC(sc10x.fmst,pc=100,hpc=0.9,file="fmst",print="2")
+sc10x.fmst <- results[[1]]
+pc.use.fmst <- results[[2]]
+rm(results)
+sc10x.fmst <- scCluster(sc10x.fmst,res=res,red="pca",dim=pc.use.st,print="2",folder="fmst")
+
+results <- scPC(sc10x.st,pc=100,hpc=0.9,file="st",print="2")
+sc10x.st <- results[[1]]
+pc.use.st <- results[[2]]
+rm(results)
+sc10x.st <- scCluster(sc10x.st,res=res,red="pca",dim=pc.use.st,print="2",folder="st")
+
+results <- scPC(sc10x.leu,pc=100,hpc=0.9,file="leu",print="2")
+sc10x.leu <- results[[1]]
+pc.use.leu <- results[[2]]
+rm(results)
+sc10x.leu <- scCluster(sc10x.leu,res=res,red="pca",dim=pc.use.leu,print="2",folder="leu.lin")
+
+
+singler.leu = CreateSinglerObject(GetAssayData(sc10x.leu,assay="RNA"), annot = NULL, "PdPbPc_deep", min.genes = 0,
+                              technology = "10X", species = "Human", citation = "",
+                              ref.list = list(cibersort), normalize.gene.length = F, variable.genes = "de",
+                              fine.tune = T, do.signatures = T, clusters = NULL, do.main.types = T, 
+                              reduce.file.size = T, numCores = SingleR.numCores)
+
+save(singler,file='./analysis/singler_object.RData')
\ No newline at end of file

r.scripts/scmatch.R

+write.csv(GetAssayData(sc10x,assay="RNA"),file=paste0("./analysis/",project.name,".csv"))

r.scripts/singler.R

@@ -2,17 +2,17 @@
 # load("./genesets/scDWS.100.rda")
 # load("./genesets/scDWS.200.rda")
 load("./genesets/scDWS.250.rda")
-load("./genesets/scDWS.lin.rda")
-load("./genesets/scDWS.epi.rda")
-load("./genesets/scDWS.fmst.rda")
+# load("./genesets/scDWS.lin.rda")
+# load("./genesets/scDWS.epi.rda")
+# load("./genesets/scDWS.fmst.rda")
 load("./genesets/cibersort.rda")
 load("./genesets/blueprint_encode.rda")
 load("./genesets/hpca.rda")
 
-singler = CreateSinglerObject(GetAssayData(object = sc10x), annot = NULL, "PdPbPc_deep", min.genes = 0,
+singler = CreateSinglerObject(GetAssayData(sc10x,assay="RNA"), annot = NULL, "PdPbPc_deep", min.genes = 0,
                               technology = "10X", species = "Human", citation = "",
-                              ref.list = list(cibersort), normalize.gene.length = F, variable.genes = "de",
-                              fine.tune = T, do.signatures = T, clusters = NULL, do.main.types = T, 
+                              ref.list = list(scDWS.250,cibersort,blueprint_encode), normalize.gene.length = F, variable.genes = "de",
+                              fine.tune = F, do.signatures = T, clusters = NULL, do.main.types = T, 
                               reduce.file.size = T, numCores = SingleR.numCores)
 
 singler$seurat <- sc10x
@@ -21,7 +21,7 @@ singler$meta.data$ALL.POPS <- sc10x$ALL.POPS
 
 singler$meta.data$xy <- sc10x[["umap"]]@cell.embeddings
 
-save(singler,file='./analysis/singler_object.RData')
+save(singler,file='./analysis/singler_testobject.RData')
 
 p <-SingleR.PlotTsne(singler$singler[[1]]$SingleR.single,singler$meta.data$xy)
 
@@ -119,3 +119,4 @@ rm(gene.set)
 
 save(sc10x,file=paste0("./analysis/",project.name,".RData"))
 save.image(file="./analysis/Data.RData")
+

r.scripts/singler.create.ref.R

 library(tidyverse)
+library(readxl)
 library(genefilter)
+library(jsonlite)
+library(SingleR)
 
 load("/work/urology/ghenry/RNA-Seq/SingleCell/PIPELINE/RUN/Pd/PUBLISH/Pd/sc-TissueMapper_Pr/analysis/sc10x.Rda")
 sc10x <- UpdateSeuratObject(sc10x)
@@ -148,9 +151,42 @@ types = as.character(colnames(cibersort))
 main_types = cibersort.main
 ref = list(name=name,data = expr, types=types, main_types=main_types)
 
-ref$de.genes = CreateVariableGeneSet(expr,types,200)
-ref$de.genes.main = CreateVariableGeneSet(expr,main_types,300)
+ref$de.genes = CreateVariableGeneSet(expr,types,300)
+ref$de.genes.main = CreateVariableGeneSet(expr,main_types,500)
 
 cibersort <- ref
 
 save(cibersort,file='./genesets/cibersort.rda')
+
+
+#fantom5 <- read.csv("./genesets/9606_symbol.csv")
+fantom5 <- read_csv("./genesets/9606_symbol.csv")
+fantom5 <- as_data_frame(as.matrix(fantom5))
+fantom5 <- column_to_rownames(fantom5, var = "symbol")
+colnames(fantom5) <-  gsub("\\..","",colnames(fantom5))
+fantom5.main.oto <- read_json("./genesets/human_samples_oto.txt",simplifyVector = TRUE)
+# fantom5.main.oto <- lapply(fantom5.main.oto,function(y) gsub(":",".",y))
+# fantom5.main.oto <- lapply(fantom5.main.oto,function(y) gsub("-",".",y))
+# fantom5.main.oto <- lapply(fantom5.main.oto,function(y) gsub(" ",".",y))
+fantom5.main.oto <- lapply(fantom5.main.oto,function(x) gsub("\\..","",x))
+fantom5 <- fantom5[,!(colnames(fantom5) %in% unlist(unname(unlist((fantom5.main.oto[grep("DOID",names(fantom5.main.oto))])))))]
+fantom5 <- fantom5[,!(colnames(fantom5) %in% unlist(unname(unlist((fantom5.main.oto[grep("UBERON",names(fantom5.main.oto))])))))]
+fantom5.main <- colnames(fantom5)
+
+
+
+name = 'fantom5'
+expr <- mapply(fantom5, FUN=as.numeric)
+expr <- matrix(data=expr,ncol=ncol(fantom5),nrow=nrow(fantom5))
+rownames(expr) <- rownames(fantom5)
+colnames(expr) <- colnames(fantom5)
+types = as.character(colnames(fantom5))
+main_types = fantom5.main
+ref = list(name=name,data = expr, types=types, main_types=main_types)
+
+ref$de.genes = CreateVariableGeneSet(expr,types,200)
+ref$de.genes.main = CreateVariableGeneSet(expr,main_types,300)
+
+fantom5 <- ref
+
+save(fantom5,file='./genesets/fantom5.rda')
\ No newline at end of file