Mercurial > repos > davidvanzessen > clonal_sequences_in_paired_samples
view RScript.r @ 28:a63ccc36f5a4 draft
Uploaded
| author | davidvanzessen |
|---|---|
| date | Tue, 19 May 2015 08:13:49 -0400 |
| parents | dd8518ea23dd |
| children | 5ab17bdf2530 |
line wrap: on
line source
args <- commandArgs(trailingOnly = TRUE) inFile = args[1] outDir = args[2] logfile = args[3] min_freq = as.numeric(args[4]) min_cells = as.numeric(args[5]) cat("<html><table><tr><td>Starting analysis</td></tr>", file=logfile, append=F) library(ggplot2) library(reshape2) library(data.table) library(grid) library(parallel) #require(xtable) cat("<tr><td>Reading input</td></tr>", file=logfile, append=T) dat = read.table(inFile, header=T, sep="\t", dec=".", fill=T, stringsAsFactors=F) dat = dat[!is.na(dat$Patient),] dat$Related_to_leukemia_clone = F setwd(outDir) cat("<tr><td>Selecting first V/J Genes</td></tr>", file=logfile, append=T) dat$V_Segment_Major_Gene = as.factor(as.character(lapply(strsplit(as.character(dat$V_Segment_Major_Gene), "; "), "[[", 1))) dat$J_Segment_Major_Gene = as.factor(as.character(lapply(strsplit(as.character(dat$J_Segment_Major_Gene), "; "), "[[", 1))) cat("<tr><td>Calculating Frequency</td></tr>", file=logfile, append=T) dat$Frequency = ((10^dat$Log10_Frequency)*100) dat = dat[dat$Frequency >= min_freq,] triplets = dat[grepl("VanDongen_cALL_14696", dat$Patient) | grepl("(16278)|(26402)|(26759)", dat$Sample),] cat("<tr><td>Normalizing to lowest cell count within locus</td></tr>", file=logfile, append=T) dat$locus_V = substring(dat$V_Segment_Major_Gene, 0, 4) dat$locus_J = substring(dat$J_Segment_Major_Gene, 0, 4) min_cell_count = data.frame(data.table(dat)[, list(min_cell_count=min(.SD$Cell_Count)), by=c("Patient", "locus_V", "locus_J")]) dat$min_cell_paste = paste(dat$Patient, dat$locus_V, dat$locus_J) min_cell_count$min_cell_paste = paste(min_cell_count$Patient, min_cell_count$locus_V, min_cell_count$locus_J) min_cell_count = min_cell_count[,c("min_cell_paste", "min_cell_count")] dat = merge(dat, min_cell_count, by="min_cell_paste") dat$normalized_read_count = round(dat$Clone_Molecule_Count_From_Spikes / dat$Cell_Count * dat$min_cell_count / 2, digits=2) #??????????????????????????????????? wel of geen / 2 dat = dat[dat$normalized_read_count >= min_cells,] dat$paste = paste(dat$Sample, dat$Clone_Sequence) patients = split(dat, dat$Patient, drop=T) intervalReads = rev(c(0,10,25,50,100,250,500,750,1000,10000)) intervalFreq = rev(c(0,0.01,0.05,0.1,0.5,1,5)) V_Segments = c(".*", "IGHV", "IGHD", "IGKV", "IGKV", "IgKINTR", "TRGV", "TRDV", "TRDD" , "TRBV") J_Segments = c(".*", ".*", ".*", "IGKJ", "KDE", ".*", ".*", ".*", ".*", ".*") Titles = c("Total", "IGH-Vh-Jh", "IGH-Dh-Jh", "Vk-Jk", "Vk-Kde" , "Intron-Kde", "TCRG", "TCRD-Vd-Dd", "TCRD-Dd-Dd", "TCRB-Vb-Jb") Titles = factor(Titles, levels=Titles) TitlesOrder = data.frame("Title"=Titles, "TitlesOrder"=1:length(Titles)) patientCountOnColumn <- function(x, product, interval, on, appendtxt=F){ if (!is.data.frame(x) & is.list(x)){ x = x[[1]] } x$Sample = factor(x$Sample, levels=unique(x$Sample)) onShort = "reads" if(on == "Frequency"){ onShort = "freq" } onx = paste(on, ".x", sep="") ony = paste(on, ".y", sep="") splt = split(x, x$Sample, drop=T) type="pair" if(length(splt) == 1){ print(paste(paste(x[1,which(colnames(x) == "Patient")]), "has one sample")) splt[[2]] = data.frame("Patient" = character(0), "Receptor" = character(0), "Sample" = character(0), "Cell_Count" = numeric(0), "Clone_Molecule_Count_From_Spikes" = numeric(0), "Log10_Frequency" = numeric(0), "Total_Read_Count" = numeric(0), "dsMol_per_1e6_cells" = numeric(0), "J_Segment_Major_Gene" = character(0), "V_Segment_Major_Gene" = character(0), "Clone_Sequence" = character(0), "CDR3_Sense_Sequence" = character(0), "Related_to_leukemia_clone" = logical(0), "Frequency"= numeric(0), "normalized_read_count" = numeric(0), "paste" = character(0)) type="single" } patient1 = splt[[1]] patient2 = splt[[2]] threshholdIndex = which(colnames(product) == "interval") V_SegmentIndex = which(colnames(product) == "V_Segments") J_SegmentIndex = which(colnames(product) == "J_Segments") titleIndex = which(colnames(product) == "Titles") sampleIndex = which(colnames(x) == "Sample") patientIndex = which(colnames(x) == "Patient") oneSample = paste(patient1[1,sampleIndex], sep="") twoSample = paste(patient2[1,sampleIndex], sep="") patient = paste(x[1,patientIndex]) switched = F if(length(grep(".*_Right$", twoSample)) == 1 || length(grep(".*_Dx_BM$", twoSample)) == 1 || length(grep(".*_Dx$", twoSample)) == 1 ){ tmp = twoSample twoSample = oneSample oneSample = tmp tmp = patient1 patient1 = patient2 patient2 = tmp switched = T } if(appendtxt){ cat(paste(patient, oneSample, twoSample, type, sep="\t"), file="patients.txt", append=T, sep="", fill=3) } cat(paste("<tr><td>", patient, "</td></tr>", sep=""), file=logfile, append=T) #patient1$merge = paste(patient1$V_Segment_Major_Gene, patient1$J_Segment_Major_Gene, patient1$CDR3_Sense_Sequence) #patient2$merge = paste(patient2$V_Segment_Major_Gene, patient2$J_Segment_Major_Gene, patient2$CDR3_Sense_Sequence) patient1$merge = paste(patient1$Clone_Sequence) patient2$merge = paste(patient2$Clone_Sequence) #patientMerge = merge(patient1, patient2, by.x="merge", by.y="merge") patientMerge = merge(patient1, patient2, by.x="merge", by.y="merge") patientMerge$thresholdValue = pmax(patientMerge[,onx], patientMerge[,ony]) res1 = vector() res2 = vector() resBoth = vector() read1Count = vector() read2Count = vector() locussum1 = vector() locussum2 = vector() print(patient) #for(iter in 1){ for(iter in 1:length(product[,1])){ threshhold = product[iter,threshholdIndex] V_Segment = paste(".*", as.character(product[iter,V_SegmentIndex]), ".*", sep="") J_Segment = paste(".*", as.character(product[iter,J_SegmentIndex]), ".*", sep="") #both = (grepl(V_Segment, patientMerge$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge$J_Segment_Major_Gene.x) & patientMerge[,onx] > threshhold & patientMerge[,ony] > threshhold) #both higher than threshold both = (grepl(V_Segment, patientMerge$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge$J_Segment_Major_Gene.x) & patientMerge$thresholdValue > threshhold) #highest of both higher than threshold one = (grepl(V_Segment, patient1$V_Segment_Major_Gene) & grepl(J_Segment, patient1$J_Segment_Major_Gene) & patient1[,on] > threshhold & !(patient1$Clone_Sequence %in% patientMerge[both,]$merge)) two = (grepl(V_Segment, patient2$V_Segment_Major_Gene) & grepl(J_Segment, patient2$J_Segment_Major_Gene) & patient2[,on] > threshhold & !(patient2$Clone_Sequence %in% patientMerge[both,]$merge)) read1Count = append(read1Count, sum(patient1[one,]$normalized_read_count)) read2Count = append(read2Count, sum(patient2[two,]$normalized_read_count)) res1 = append(res1, sum(one)) res2 = append(res2, sum(two)) resBoth = append(resBoth, sum(both)) locussum1 = append(locussum1, sum(patient1[(grepl(V_Segment, patient1$V_Segment_Major_Gene) & grepl(J_Segment, patient1$J_Segment_Major_Gene)),]$normalized_read_count)) locussum2 = append(locussum2, sum(patient2[(grepl(V_Segment, patient2$V_Segment_Major_Gene) & grepl(J_Segment, patient2$J_Segment_Major_Gene)),]$normalized_read_count)) #threshhold = 0 if(threshhold != 0){ if(sum(one) > 0){ dfOne = patient1[one,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfOne) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone Sequence", "Related_to_leukemia_clone") filenameOne = paste(oneSample, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfOne, file=paste(filenameOne, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(two) > 0){ dfTwo = patient2[two,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfTwo) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone Sequence", "Related_to_leukemia_clone") filenameTwo = paste(twoSample, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfTwo, file=paste(filenameTwo, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } } if(sum(both) > 0){ dfBoth = patientMerge[both,c("V_Segment_Major_Gene.x", "J_Segment_Major_Gene.x", "normalized_read_count.x", "Frequency.x", "Related_to_leukemia_clone.x", "Clone_Sequence.x", "V_Segment_Major_Gene.y", "J_Segment_Major_Gene.y", "normalized_read_count.y", "Frequency.y", "Related_to_leukemia_clone.y")] colnames(dfBoth) = c(paste("Proximal segment", oneSample), paste("Distal segment", oneSample), paste("Normalized_Read_Count", oneSample), paste("Frequency", oneSample), paste("Related_to_leukemia_clone", oneSample),"Clone Sequence", paste("Proximal segment", twoSample), paste("Distal segment", twoSample), paste("Normalized_Read_Count", twoSample), paste("Frequency", twoSample), paste("Related_to_leukemia_clone", twoSample)) filenameBoth = paste(oneSample, "_", twoSample, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfBoth, file=paste(filenameBoth, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } } patientResult = data.frame("Locus"=product$Titles, "J_Segment"=product$J_Segments, "V_Segment"=product$V_Segments, "cut_off_value"=paste(">", product$interval, sep=""), "Both"=resBoth, "tmp1"=res1, "read_count1" = round(read1Count), "tmp2"=res2, "read_count2"= round(read2Count), "Sum"=res1 + res2 + resBoth, "percentage" = round((resBoth/(res1 + res2 + resBoth)) * 100, digits=2), "Locus_sum1"=locussum1, "Locus_sum2"=locussum2) if(sum(is.na(patientResult$percentage)) > 0){ patientResult[is.na(patientResult$percentage),]$percentage = 0 } colnames(patientResult)[6] = oneSample colnames(patientResult)[8] = twoSample colnamesBak = colnames(patientResult) colnames(patientResult) = c("Ig/TCR gene rearrangement type", "Distal Gene segment", "Proximal gene segment", "cut_off_value", paste("Number of sequences ", patient, "_Both", sep=""), paste("Number of sequences", oneSample, sep=""), paste("Normalized Read Count", oneSample), paste("Number of sequences", twoSample, sep=""), paste("Normalized Read Count", twoSample), paste("Sum number of sequences", patient), paste("Percentage of sequences ", patient, "_Both", sep=""), paste("Locus Sum", oneSample), paste("Locus Sum", twoSample)) write.table(patientResult, file=paste(patient, "_", onShort, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) colnames(patientResult) = colnamesBak patientResult$Locus = factor(patientResult$Locus, Titles) patientResult$cut_off_value = factor(patientResult$cut_off_value, paste(">", interval, sep="")) plt = ggplot(patientResult[,c("Locus", "cut_off_value", "Both")]) plt = plt + geom_bar( aes( x=factor(cut_off_value), y=Both), stat='identity', position="dodge", fill="#79c36a") plt = plt + facet_grid(.~Locus) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) plt = plt + geom_text(aes(ymax=max(Both), x=cut_off_value,y=Both,label=Both), angle=90, hjust=0) plt = plt + xlab("Reads per locus") + ylab("Count") + ggtitle("Number of clones in both") plt = plt + theme(plot.margin = unit(c(1,8.8,0.5,1.5), "lines")) png(paste(patient, "_", onShort, ".png", sep=""), width=1920, height=1080) print(plt) dev.off() #(t,r,b,l) plt = ggplot(patientResult[,c("Locus", "cut_off_value", "percentage")]) plt = plt + geom_bar( aes( x=factor(cut_off_value), y=percentage), stat='identity', position="dodge", fill="#79c36a") plt = plt + facet_grid(.~Locus) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) plt = plt + geom_text(aes(ymax=max(percentage), x=cut_off_value,y=percentage,label=percentage), angle=90, hjust=0) plt = plt + xlab("Reads per locus") + ylab("Count") + ggtitle("% clones in both left and right") plt = plt + theme(plot.margin = unit(c(1,8.8,0.5,1.5), "lines")) png(paste(patient, "_percent_", onShort, ".png", sep=""), width=1920, height=1080) print(plt) dev.off() patientResult = melt(patientResult[,c('Locus','cut_off_value', oneSample, twoSample)] ,id.vars=1:2) patientResult$relativeValue = patientResult$value * 10 patientResult[patientResult$relativeValue == 0,]$relativeValue = 1 plt = ggplot(patientResult) plt = plt + geom_bar( aes( x=factor(cut_off_value), y=relativeValue, fill=variable), stat='identity', position="dodge") plt = plt + facet_grid(.~Locus) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) plt = plt + scale_y_continuous(trans="log", breaks=10^c(0:10), labels=c(0, 10^c(0:9))) plt = plt + geom_text(data=patientResult[patientResult$variable == oneSample,], aes(ymax=max(value), x=cut_off_value,y=relativeValue,label=value), angle=90, position=position_dodge(width=0.9), hjust=0, vjust=-0.2) plt = plt + geom_text(data=patientResult[patientResult$variable == twoSample,], aes(ymax=max(value), x=cut_off_value,y=relativeValue,label=value), angle=90, position=position_dodge(width=0.9), hjust=0, vjust=0.8) plt = plt + xlab("Reads per locus") + ylab("Count") + ggtitle(paste("Number of clones in only ", oneSample, " and only ", twoSample, sep="")) png(paste(patient, "_", onShort, "_both.png", sep=""), width=1920, height=1080) print(plt) dev.off() } cat("<tr><td>Starting Frequency analysis</td></tr>", file=logfile, append=T) interval = intervalFreq intervalOrder = data.frame("interval"=paste(">", interval, sep=""), "intervalOrder"=1:length(interval)) product = data.frame("Titles"=rep(Titles, each=length(interval)), "interval"=rep(interval, times=10), "V_Segments"=rep(V_Segments, each=length(interval)), "J_Segments"=rep(J_Segments, each=length(interval))) mclapply(patients, FUN=patientCountOnColumn, product = product, interval=interval, on="Frequency", appendtxt=T) cat("<tr><td>Starting Cell Count analysis</td></tr>", file=logfile, append=T) interval = intervalReads intervalOrder = data.frame("interval"=paste(">", interval, sep=""), "intervalOrder"=1:length(interval)) product = data.frame("Titles"=rep(Titles, each=length(interval)), "interval"=rep(interval, times=10), "V_Segments"=rep(V_Segments, each=length(interval)), "J_Segments"=rep(J_Segments, each=length(interval))) mclapply(patients, FUN=patientCountOnColumn, product = product, interval=interval, on="normalized_read_count") cat("</table></html>", file=logfile, append=T) tripletAnalysis <- function(patient1, label1, patient2, label2, patient3, label3, product, interval, on, appendTriplets= FALSE){ onShort = "reads" if(on == "Frequency"){ onShort = "freq" } onx = paste(on, ".x", sep="") ony = paste(on, ".y", sep="") onz = paste(on, ".z", sep="") type="triplet" threshholdIndex = which(colnames(product) == "interval") V_SegmentIndex = which(colnames(product) == "V_Segments") J_SegmentIndex = which(colnames(product) == "J_Segments") titleIndex = which(colnames(product) == "Titles") sampleIndex = which(colnames(patient1) == "Sample") patientIndex = which(colnames(patient1) == "Patient") oneSample = paste(patient1[1,sampleIndex], sep="") twoSample = paste(patient2[1,sampleIndex], sep="") threeSample = paste(patient3[1,sampleIndex], sep="") #patient1$merge = paste(patient1$V_Segment_Major_Gene, patient1$J_Segment_Major_Gene, patient1$CDR3_Sense_Sequence) #patient2$merge = paste(patient2$V_Segment_Major_Gene, patient2$J_Segment_Major_Gene, patient2$CDR3_Sense_Sequence) #patient3$merge = paste(patient3$V_Segment_Major_Gene, patient3$J_Segment_Major_Gene, patient3$CDR3_Sense_Sequence) patient1$merge = paste(patient1$Clone_Sequence) patient2$merge = paste(patient2$Clone_Sequence) patient3$merge = paste(patient3$Clone_Sequence) patientMerge = merge(patient1, patient2, by="merge") patientMerge = merge(patientMerge, patient3, by="merge") colnames(patientMerge)[which(!grepl("(\\.x$)|(\\.y$)|(merge)", names(patientMerge)))] = paste(colnames(patientMerge)[which(!grepl("(\\.x$)|(\\.y$)|(merge)", names(patientMerge), perl=T))], ".z", sep="") patientMerge$thresholdValue = pmax(patientMerge[,onx], patientMerge[,ony], patientMerge[,onz]) patientMerge12 = merge(patient1, patient2, by="merge") patientMerge12$thresholdValue = pmax(patientMerge12[,onx], patientMerge12[,ony]) patientMerge13 = merge(patient1, patient3, by="merge") patientMerge13$thresholdValue = pmax(patientMerge13[,onx], patientMerge13[,ony]) patientMerge23 = merge(patient2, patient3, by="merge") patientMerge23$thresholdValue = pmax(patientMerge23[,onx], patientMerge23[,ony]) scatterplot_data_columns = c("Clone_Sequence", "Frequency", "normalized_read_count", "V_Segment_Major_Gene", "J_Segment_Major_Gene") scatterplot_data = rbind(patient1[,scatterplot_data_columns], patient2[,scatterplot_data_columns], patient3[,scatterplot_data_columns]) scatterplot_data = scatterplot_data[!duplicated(scatterplot_data$Clone_Sequence),] scatterplot_data$type = factor(x="In one", levels=c("In one", "In two", "In three", "In multiple")) res1 = vector() res2 = vector() res3 = vector() res12 = vector() res13 = vector() res23 = vector() resAll = vector() read1Count = vector() read2Count = vector() read3Count = vector() if(appendTriplets){ cat(paste(label1, label2, label3, sep="\t"), file="triplets.txt", append=T, sep="", fill=3) } for(iter in 1:length(product[,1])){ threshhold = product[iter,threshholdIndex] V_Segment = paste(".*", as.character(product[iter,V_SegmentIndex]), ".*", sep="") J_Segment = paste(".*", as.character(product[iter,J_SegmentIndex]), ".*", sep="") #all = (grepl(V_Segment, patientMerge$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge$J_Segment_Major_Gene.x) & patientMerge[,onx] > threshhold & patientMerge[,ony] > threshhold & patientMerge[,onz] > threshhold) all = (grepl(V_Segment, patientMerge$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge$J_Segment_Major_Gene.x) & patientMerge$thresholdValue > threshhold) one_two = (grepl(V_Segment, patientMerge12$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge12$J_Segment_Major_Gene.x) & patientMerge12$thresholdValue > threshhold & !(patientMerge12$Clone_Sequence.x %in% patientMerge[all,]$merge)) one_three = (grepl(V_Segment, patientMerge13$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge13$J_Segment_Major_Gene.x) & patientMerge13$thresholdValue > threshhold & !(patientMerge13$Clone_Sequence.x %in% patientMerge[all,]$merge)) two_three = (grepl(V_Segment, patientMerge23$V_Segment_Major_Gene.x) & grepl(J_Segment, patientMerge23$J_Segment_Major_Gene.x) & patientMerge23$thresholdValue > threshhold & !(patientMerge23$Clone_Sequence.x %in% patientMerge[all,]$merge)) one = (grepl(V_Segment, patient1$V_Segment_Major_Gene) & grepl(J_Segment, patient1$J_Segment_Major_Gene) & patient1[,on] > threshhold & !(patient1$Clone_Sequence %in% patientMerge[all,]$merge) & !(patient1$Clone_Sequence %in% patientMerge12[one_two,]$merge) & !(patient1$Clone_Sequence %in% patientMerge13[one_three,]$merge)) two = (grepl(V_Segment, patient2$V_Segment_Major_Gene) & grepl(J_Segment, patient2$J_Segment_Major_Gene) & patient2[,on] > threshhold & !(patient2$Clone_Sequence %in% patientMerge[all,]$merge) & !(patient2$Clone_Sequence %in% patientMerge12[one_two,]$merge) & !(patient2$Clone_Sequence %in% patientMerge23[two_three,]$merge)) three = (grepl(V_Segment, patient3$V_Segment_Major_Gene) & grepl(J_Segment, patient3$J_Segment_Major_Gene) & patient3[,on] > threshhold & !(patient3$Clone_Sequence %in% patientMerge[all,]$merge) & !(patient3$Clone_Sequence %in% patientMerge13[one_three,]$merge) & !(patient3$Clone_Sequence %in% patientMerge23[two_three,]$merge)) read1Count = append(read1Count, sum(patient1[one,]$normalized_read_count) + sum(patientMerge[all,]$normalized_read_count.x)) read2Count = append(read2Count, sum(patient2[two,]$normalized_read_count) + sum(patientMerge[all,]$normalized_read_count.y)) read3Count = append(read3Count, sum(patient3[three,]$normalized_read_count) + sum(patientMerge[all,]$normalized_read_count.z)) res1 = append(res1, sum(one)) res2 = append(res2, sum(two)) res3 = append(res3, sum(three)) resAll = append(resAll, sum(all)) res12 = append(res12, sum(one_two)) res13 = append(res13, sum(one_three)) res23 = append(res23, sum(two_three)) #threshhold = 0 if(threshhold != 0){ if(sum(one) > 0){ dfOne = patient1[one,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfOne) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone") filenameOne = paste(label1, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfOne, file=paste(filenameOne, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(two) > 0){ dfTwo = patient2[two,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfTwo) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone") filenameTwo = paste(label2, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfTwo, file=paste(filenameTwo, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(three) > 0){ dfThree = patient3[three,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfThree) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone") filenameThree = paste(label3, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfThree, file=paste(filenameThree, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(one_two) > 0){ dfOne_two = patientMerge12[one_two,c("V_Segment_Major_Gene.x", "J_Segment_Major_Gene.x", "normalized_read_count.x", "Frequency.x", "Related_to_leukemia_clone.x", "Clone_Sequence.x", "V_Segment_Major_Gene.y", "J_Segment_Major_Gene.y", "normalized_read_count.y", "Frequency.y", "Related_to_leukemia_clone.y")] colnames(dfOne_two) = c(paste("Proximal segment", oneSample), paste("Distal segment", oneSample), paste("Normalized_Read_Count", oneSample), paste("Frequency", oneSample), paste("Related_to_leukemia_clone", oneSample),"Clone_Sequence", paste("Proximal segment", twoSample), paste("Distal segment", twoSample), paste("Normalized_Read_Count", twoSample), paste("Frequency", twoSample), paste("Related_to_leukemia_clone", twoSample)) filenameOne_two = paste(label1, "_", label2, "_", product[iter, titleIndex], "_", threshhold, onShort, sep="") write.table(dfOne_two, file=paste(filenameOne_two, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(one_three) > 0){ dfOne_three = patientMerge13[one_three,c("V_Segment_Major_Gene.x", "J_Segment_Major_Gene.x", "normalized_read_count.x", "Frequency.x", "Related_to_leukemia_clone.x", "Clone_Sequence.x", "V_Segment_Major_Gene.y", "J_Segment_Major_Gene.y", "normalized_read_count.y", "Frequency.y", "Related_to_leukemia_clone.y")] colnames(dfOne_three) = c(paste("Proximal segment", oneSample), paste("Distal segment", oneSample), paste("Normalized_Read_Count", oneSample), paste("Frequency", oneSample), paste("Related_to_leukemia_clone", oneSample),"Clone_Sequence", paste("Proximal segment", threeSample), paste("Distal segment", threeSample), paste("Normalized_Read_Count", threeSample), paste("Frequency", threeSample), paste("Related_to_leukemia_clone", threeSample)) filenameOne_three = paste(label1, "_", label3, "_", product[iter, titleIndex], "_", threshhold, onShort, sep="") write.table(dfOne_three, file=paste(filenameOne_three, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(two_three) > 0){ dfTwo_three = patientMerge23[two_three,c("V_Segment_Major_Gene.x", "J_Segment_Major_Gene.x", "normalized_read_count.x", "Frequency.x", "Related_to_leukemia_clone.x", "Clone_Sequence.x", "V_Segment_Major_Gene.y", "J_Segment_Major_Gene.y", "normalized_read_count.y", "Frequency.y", "Related_to_leukemia_clone.y")] colnames(dfTwo_three) = c(paste("Proximal segment", twoSample), paste("Distal segment", twoSample), paste("Normalized_Read_Count", twoSample), paste("Frequency", twoSample), paste("Related_to_leukemia_clone", twoSample),"Clone_Sequence", paste("Proximal segment", threeSample), paste("Distal segment", threeSample), paste("Normalized_Read_Count", threeSample), paste("Frequency", threeSample), paste("Related_to_leukemia_clone", threeSample)) filenameTwo_three = paste(label2, "_", label3, "_", product[iter, titleIndex], "_", threshhold, onShort, sep="") write.table(dfTwo_three, file=paste(filenameTwo_three, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } } else { #scatterplot data scatterplot_locus_data = scatterplot_data[grepl(V_Segment, scatterplot_data$V_Segment_Major_Gene) & grepl(J_Segment, scatterplot_data$J_Segment_Major_Gene),] in_two = (scatterplot_locus_data$Clone_Sequence %in% patientMerge12[one_two,]$Clone_Sequence.x) | (scatterplot_locus_data$Clone_Sequence %in% patientMerge13[one_three,]$Clone_Sequence.x) | (scatterplot_locus_data$Clone_Sequence %in% patientMerge23[two_three,]$Clone_Sequence.x) if(sum(in_two) > 0){ scatterplot_locus_data[in_two,]$type = "In two" } in_three = (scatterplot_locus_data$Clone_Sequence %in% patientMerge[all,]$Clone_Sequence.x) if(sum(in_three)> 0){ scatterplot_locus_data[in_three,]$type = "In three" } not_in_one = scatterplot_locus_data$type != "In one" if(sum(not_in_one) > 0){ scatterplot_locus_data[not_in_one,]$type = "In multiple" } p = NULL if(nrow(scatterplot_locus_data) != 0){ if(on == "normalized_read_count"){ scales = 10^(0:ceiling(log10(max(scatterplot_locus_data$normalized_read_count)))) p = ggplot(scatterplot_locus_data, aes(type, normalized_read_count)) + scale_y_log10(breaks=scales,labels=scales) } else { p = ggplot(scatterplot_locus_data, aes(type, Frequency)) } p = p + geom_point(aes(colour=type), position="jitter") p = p + xlab("In one or in multiple samples") + ylab(onShort) + ggtitle(paste(label1, label2, label3, onShort, product[iter, titleIndex])) } else { p = ggplot(NULL, aes(x=c("In one", "In multiple"),y=0)) + geom_blank(NULL) + xlab("In two or in three of the samples") + ylab(onShort) + ggtitle(paste(label1, label2, label3, onShort, product[iter, titleIndex])) } png(paste(label1, "_", label2, "_", label3, "_", onShort, "_", product[iter, titleIndex],"_scatter.png", sep="")) print(p) dev.off() } if(sum(all) > 0){ dfAll = patientMerge[all,c("V_Segment_Major_Gene.x", "J_Segment_Major_Gene.x", "normalized_read_count.x", "Frequency.x", "Related_to_leukemia_clone.x", "Clone_Sequence.x", "V_Segment_Major_Gene.y", "J_Segment_Major_Gene.y", "normalized_read_count.y", "Frequency.y", "Related_to_leukemia_clone.y", "V_Segment_Major_Gene.z", "J_Segment_Major_Gene.z", "normalized_read_count.z", "Frequency.z", "Related_to_leukemia_clone.z")] colnames(dfAll) = c(paste("Proximal segment", oneSample), paste("Distal segment", oneSample), paste("Normalized_Read_Count", oneSample), paste("Frequency", oneSample), paste("Related_to_leukemia_clone", oneSample),"Clone_Sequence", paste("Proximal segment", twoSample), paste("Distal segment", twoSample), paste("Normalized_Read_Count", twoSample), paste("Frequency", twoSample), paste("Related_to_leukemia_clone", twoSample), paste("Proximal segment", threeSample), paste("Distal segment", threeSample), paste("Normalized_Read_Count", threeSample), paste("Frequency", threeSample), paste("Related_to_leukemia_clone", threeSample)) filenameAll = paste(label1, "_", label2, "_", label3, "_", product[iter, titleIndex], "_", threshhold, sep="") write.table(dfAll, file=paste(filenameAll, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } } #patientResult = data.frame("Locus"=product$Titles, "J_Segment"=product$J_Segments, "V_Segment"=product$V_Segments, "cut_off_value"=paste(">", product$interval, sep=""), "All"=resAll, "tmp1"=res1, "read_count1" = round(read1Count), "tmp2"=res2, "read_count2"= round(read2Count), "tmp3"=res3, "read_count3"=round(read3Count)) patientResult = data.frame("Locus"=product$Titles, "J_Segment"=product$J_Segments, "V_Segment"=product$V_Segments, "cut_off_value"=paste(">", product$interval, sep=""), "All"=resAll, "tmp1"=res1, "tmp2"=res2, "tmp3"=res3, "tmp12"=res12, "tmp13"=res13, "tmp23"=res23) colnames(patientResult)[6] = oneSample colnames(patientResult)[7] = twoSample colnames(patientResult)[8] = threeSample colnames(patientResult)[9] = paste(oneSample, twoSample, sep="_") colnames(patientResult)[10] = paste(oneSample, twoSample, sep="_") colnames(patientResult)[11] = paste(oneSample, twoSample, sep="_") colnamesBak = colnames(patientResult) colnames(patientResult) = c("Ig/TCR gene rearrangement type", "Distal Gene segment", "Proximal gene segment", "cut_off_value", "Number of sequences All", paste("Number of sequences", oneSample), paste("Number of sequences", twoSample), paste("Number of sequences", threeSample), paste("Number of sequences", oneSample, twoSample), paste("Number of sequences", oneSample, threeSample), paste("Number of sequences", twoSample, threeSample)) write.table(patientResult, file=paste(label1, "_", label2, "_", label3, "_", onShort, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) colnames(patientResult) = colnamesBak patientResult$Locus = factor(patientResult$Locus, Titles) patientResult$cut_off_value = factor(patientResult$cut_off_value, paste(">", interval, sep="")) plt = ggplot(patientResult[,c("Locus", "cut_off_value", "All")]) plt = plt + geom_bar( aes( x=factor(cut_off_value), y=All), stat='identity', position="dodge", fill="#79c36a") plt = plt + facet_grid(.~Locus) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) plt = plt + geom_text(aes(ymax=max(All), x=cut_off_value,y=All,label=All), angle=90, hjust=0) plt = plt + xlab("Reads per locus") + ylab("Count") + ggtitle("Number of clones in All") plt = plt + theme(plot.margin = unit(c(1,8.8,0.5,1.5), "lines")) png(paste(label1, "_", label2, "_", label3, "_", onShort, "_total_all.png", sep=""), width=1920, height=1080) print(plt) dev.off() fontSize = 4 bak = patientResult patientResult = melt(patientResult[,c('Locus','cut_off_value', oneSample, twoSample, threeSample)] ,id.vars=1:2) patientResult$relativeValue = patientResult$value * 10 patientResult[patientResult$relativeValue == 0,]$relativeValue = 1 plt = ggplot(patientResult) plt = plt + geom_bar( aes( x=factor(cut_off_value), y=relativeValue, fill=variable), stat='identity', position="dodge") plt = plt + facet_grid(.~Locus) + theme(axis.text.x = element_text(angle = 45, hjust = 1)) plt = plt + scale_y_continuous(trans="log", breaks=10^c(0:10), labels=c(0, 10^c(0:9))) plt = plt + geom_text(data=patientResult[patientResult$variable == oneSample,], aes(ymax=max(value), x=cut_off_value,y=relativeValue,label=value), angle=90, position=position_dodge(width=0.9), hjust=0, vjust=-0.7, size=fontSize) plt = plt + geom_text(data=patientResult[patientResult$variable == twoSample,], aes(ymax=max(value), x=cut_off_value,y=relativeValue,label=value), angle=90, position=position_dodge(width=0.9), hjust=0, vjust=0.4, size=fontSize) plt = plt + geom_text(data=patientResult[patientResult$variable == threeSample,], aes(ymax=max(value), x=cut_off_value,y=relativeValue,label=value), angle=90, position=position_dodge(width=0.9), hjust=0, vjust=1.5, size=fontSize) plt = plt + xlab("Reads per locus") + ylab("Count") + ggtitle("Number of clones in only one sample") png(paste(label1, "_", label2, "_", label3, "_", onShort, "_indiv_all.png", sep=""), width=1920, height=1080) print(plt) dev.off() } triplets$uniqueID = "ID" triplets[grepl("16278_Left", triplets$Sample),]$uniqueID = "16278_26402_26759_Left" triplets[grepl("26402_Left", triplets$Sample),]$uniqueID = "16278_26402_26759_Left" triplets[grepl("26759_Left", triplets$Sample),]$uniqueID = "16278_26402_26759_Left" triplets[grepl("16278_Right", triplets$Sample),]$uniqueID = "16278_26402_26759_Right" triplets[grepl("26402_Right", triplets$Sample),]$uniqueID = "16278_26402_26759_Right" triplets[grepl("26759_Right", triplets$Sample),]$uniqueID = "16278_26402_26759_Right" triplets[grepl("14696", triplets$Patient),]$uniqueID = "14696" triplets$locus_V = substring(triplets$V_Segment_Major_Gene, 0, 4) triplets$locus_J = substring(triplets$J_Segment_Major_Gene, 0, 4) min_cell_count = data.frame(data.table(triplets)[, list(min_cell_count=min(.SD$Cell_Count)), by=c("uniqueID", "locus_V", "locus_J")]) triplets$min_cell_paste = paste(triplets$uniqueID, triplets$locus_V, triplets$locus_J) min_cell_count$min_cell_paste = paste(min_cell_count$uniqueID, min_cell_count$locus_V, min_cell_count$locus_J) min_cell_count = min_cell_count[,c("min_cell_paste", "min_cell_count")] triplets = merge(triplets, min_cell_count, by="min_cell_paste") triplets$normalized_read_count = round(triplets$Clone_Molecule_Count_From_Spikes / triplets$Cell_Count * triplets$min_cell_count / 2, digits=2) #??????????????????????????????????? wel of geen / 2 triplets = triplets[triplets$normalized_read_count >= min_cells,] column_drops = c("locus_V", "locus_J", "min_cell_count", "min_cell_paste") triplets = triplets[,!(colnames(triplets) %in% column_drops)] interval = intervalReads intervalOrder = data.frame("interval"=paste(">", interval, sep=""), "intervalOrder"=1:length(interval)) product = data.frame("Titles"=rep(Titles, each=length(interval)), "interval"=rep(interval, times=10), "V_Segments"=rep(V_Segments, each=length(interval)), "J_Segments"=rep(J_Segments, each=length(interval))) one = triplets[triplets$Sample == "14696_reg_BM",] two = triplets[triplets$Sample == "24536_reg_BM",] three = triplets[triplets$Sample == "24062_reg_BM",] tripletAnalysis(one, "14696_1", two, "14696_2", three, "14696_3", product=product, interval=interval, on="normalized_read_count", T) one = triplets[triplets$Sample == "16278_Left",] two = triplets[triplets$Sample == "26402_Left",] three = triplets[triplets$Sample == "26759_Left",] tripletAnalysis(one, "16278_Left", two, "26402_Left", three, "26759_Left", product=product, interval=interval, on="normalized_read_count", T) one = triplets[triplets$Sample == "16278_Right",] two = triplets[triplets$Sample == "26402_Right",] three = triplets[triplets$Sample == "26759_Right",] tripletAnalysis(one, "16278_Right", two, "26402_Right", three, "26759_Right", product=product, interval=interval, on="normalized_read_count", T) interval = intervalFreq intervalOrder = data.frame("interval"=paste(">", interval, sep=""), "intervalOrder"=1:length(interval)) product = data.frame("Titles"=rep(Titles, each=length(interval)), "interval"=rep(interval, times=10), "V_Segments"=rep(V_Segments, each=length(interval)), "J_Segments"=rep(J_Segments, each=length(interval))) one = triplets[triplets$Sample == "14696_reg_BM",] two = triplets[triplets$Sample == "24536_reg_BM",] three = triplets[triplets$Sample == "24062_reg_BM",] tripletAnalysis(one, "14696_1", two, "14696_2", three, "14696_3", product=product, interval=interval, on="Frequency", F) one = triplets[triplets$Sample == "16278_Left",] two = triplets[triplets$Sample == "26402_Left",] three = triplets[triplets$Sample == "26759_Left",] tripletAnalysis(one, "16278_Left", two, "26402_Left", three, "26759_Left", product=product, interval=interval, on="Frequency", F) one = triplets[triplets$Sample == "16278_Right",] two = triplets[triplets$Sample == "26402_Right",] three = triplets[triplets$Sample == "26759_Right",] tripletAnalysis(one, "16278_Right", two, "26402_Right", three, "26759_Right", product=product, interval=interval, on="Frequency", F)