Mercurial > repos > davidvanzessen > prisca
diff RScript.r @ 2:7ffd0fba8cf4 draft
Uploaded
| author | davidvanzessen |
|---|---|
| date | Mon, 18 Sep 2017 09:01:19 -0400 |
| parents | 75853bceec00 |
| children | 20f0df3721aa |
line wrap: on
line diff
--- a/RScript.r Tue Jan 17 07:24:44 2017 -0500 +++ b/RScript.r Mon Sep 18 09:01:19 2017 -0400 @@ -18,7 +18,23 @@ #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[,c("Patient", "Receptor", "Sample", "Cell_Count", "Clone_Molecule_Count_From_Spikes", "Log10_Frequency", "Total_Read_Count", "J_Segment_Major_Gene", "V_Segment_Major_Gene", "CDR3_Sense_Sequence", "Clone_Sequence")] + +needed_cols = c("Patient", "Receptor", "Sample", "Cell_Count", "Clone_Molecule_Count_From_Spikes", "Log10_Frequency", "J_Segment_Major_Gene", "V_Segment_Major_Gene", "CDR3_Sense_Sequence", "Clone_Sequence") +if(!all(needed_cols %in% names(dat))){ + cat("Missing column(s):<br />", file=logfile, append=F) + missing_cols = needed_cols[!(needed_cols %in% names(dat))] + for(col in missing_cols){ + cat(paste(col, "<br />"), file=logfile, append=T) + } + stop("Not all columns are present") +} + +if(!("Total_Read_Count" %in% names(dat))){ + dat$Total_Read_Count = 0 +} + +dat = dat[,c("Patient", "Receptor", "Sample", "Cell_Count", "Clone_Molecule_Count_From_Spikes", "Log10_Frequency", "Total_Read_Count", "J_Segment_Major_Gene", "V_Segment_Major_Gene", "CDR3_Sense_Sequence", "Clone_Sequence")] + dat$dsPerM = 0 dat = dat[!is.na(dat$Patient),] dat$Related_to_leukemia_clone = F @@ -76,7 +92,7 @@ Titles = factor(Titles, levels=Titles) TitlesOrder = data.frame("Title"=Titles, "TitlesOrder"=1:length(Titles)) -single_patients = data.frame("Patient" = character(0),"Sample" = character(0), "on" = character(0), "Clone_Sequence" = character(0), "Frequency" = numeric(0), "normalized_read_count" = numeric(0), "V_Segment_Major_Gene" = character(0), "J_Segment_Major_Gene" = character(0), "Rearrangement" = character(0)) +single_patients = dat[NULL,] patient.merge.list = list() #cache the 'both' table, 2x speedup for more memory... patient.merge.list.second = list() @@ -96,25 +112,20 @@ onx = paste(on, ".x", sep="") ony = paste(on, ".y", sep="") splt = split(x, x$Sample, drop=T) + print(splt) type="pair" if(length(splt) == 1){ print(paste(paste(x[1,which(colnames(x) == "Patient")]), "has one sample")) - splt[[2]] = data.frame("Patient" = character(0), "Receptor" = character(0), "Sample" = character(0), "Cell_Count" = numeric(0), "Clone_Molecule_Count_From_Spikes" = numeric(0), "Log10_Frequency" = numeric(0), "Total_Read_Count" = numeric(0), "dsMol_per_1e6_cells" = numeric(0), "J_Segment_Major_Gene" = character(0), "V_Segment_Major_Gene" = character(0), "Clone_Sequence" = character(0), "CDR3_Sense_Sequence" = character(0), "Related_to_leukemia_clone" = logical(0), "Frequency"= numeric(0), "normalized_read_count" = numeric(0), "paste" = character(0)) + splt[[2]] = splt[[1]][NULL,] type="single" } patient1 = splt[[1]] patient2 = splt[[2]] - threshholdIndex = which(colnames(product) == "interval") - V_SegmentIndex = which(colnames(product) == "V_Segments") - J_SegmentIndex = which(colnames(product) == "J_Segments") - titleIndex = which(colnames(product) == "Titles") - sampleIndex = which(colnames(x) == "Sample") - patientIndex = which(colnames(x) == "Patient") - oneSample = paste(patient1[1,sampleIndex], sep="") - twoSample = paste(patient2[1,sampleIndex], sep="") - patient = paste(x[1,patientIndex]) - + oneSample = patient1[1,"Sample"] + twoSample = patient2[1,"Sample"] + patient = x[1,"Patient"] + switched = F if(length(grep(".*_Right$", twoSample)) == 1 || length(grep(".*_Dx_BM$", twoSample)) == 1 || length(grep(".*_Dx$", twoSample)) == 1 ){ tmp = twoSample @@ -139,95 +150,74 @@ } scatterplot_data_columns = c("Patient", "Sample", "Frequency", "normalized_read_count", "V_Segment_Major_Gene", "J_Segment_Major_Gene", "merge") - #scatterplot_data = rbind(patient1[,scatterplot_data_columns], patient2[,scatterplot_data_columns]) scatterplot_data = patient1[NULL,scatterplot_data_columns] - #scatterplot_data = scatterplot_data[!duplicated(scatterplot_data$merge),] - #scatterplot_data$type = factor(x=oneSample, levels=c(oneSample, twoSample, "In Both")) scatterplot.data.type.factor = c(oneSample, twoSample, paste(c(oneSample, twoSample), "In Both")) - #scatterplot_data$type = factor(x=NULL, levels=scatterplot.data.type.factor) scatterplot_data$type = character(0) scatterplot_data$link = numeric(0) scatterplot_data$on = character(0) - #patientMerge = merge(patient1, patient2, by.x="merge", by.y="merge") #merge alles 'fuzzy' patientMerge = merge(patient1, patient2, by.x="merge", by.y="merge")[NULL,] #blegh - + cs.exact.matches = patient1[patient1$Clone_Sequence %in% patient2$Clone_Sequence,]$Clone_Sequence - + start.time = proc.time() merge.list = c() - + if(patient %in% names(patient.merge.list)){ patientMerge = patient.merge.list[[patient]] merge.list[["second"]] = patient.merge.list.second[[patient]] scatterplot_data = scatter_locus_data_list[[patient]] cat(paste("<td>", nrow(patient1), " in ", oneSample, " and ", nrow(patient2), " in ", twoSample, ", ", nrow(patientMerge), " in both (fetched from cache)</td></tr>", sep=""), file=logfile, append=T) - + print(names(patient.merge.list)) } else { #fuzzy matching here... - #merge.list = patientMerge$merge - - #patient1.fuzzy = patient1[!(patient1$merge %in% merge.list),] - #patient2.fuzzy = patient2[!(patient2$merge %in% merge.list),] - + patient1.fuzzy = patient1 patient2.fuzzy = patient2 - - #patient1.fuzzy$merge = paste(patient1.fuzzy$V_Segment_Major_Gene, patient1.fuzzy$J_Segment_Major_Gene, patient1.fuzzy$CDR3_Sense_Sequence) - #patient2.fuzzy$merge = paste(patient2.fuzzy$V_Segment_Major_Gene, patient2.fuzzy$J_Segment_Major_Gene, patient2.fuzzy$CDR3_Sense_Sequence) - - #patient1.fuzzy$merge = paste(patient1.fuzzy$locus_V, patient1.fuzzy$locus_J, patient1.fuzzy$CDR3_Sense_Sequence) - #patient2.fuzzy$merge = paste(patient2.fuzzy$locus_V, patient2.fuzzy$locus_J, patient2.fuzzy$CDR3_Sense_Sequence) - + patient1.fuzzy$merge = paste(patient1.fuzzy$locus_V, patient1.fuzzy$locus_J) patient2.fuzzy$merge = paste(patient2.fuzzy$locus_V, patient2.fuzzy$locus_J) - - #merge.freq.table = data.frame(table(c(patient1.fuzzy[!duplicated(patient1.fuzzy$merge),"merge"], patient2.fuzzy[!duplicated(patient2.fuzzy$merge),"merge"]))) #also remove? - #merge.freq.table.gt.1 = merge.freq.table[merge.freq.table$Freq > 1,] - - #patient1.fuzzy = patient1.fuzzy[patient1.fuzzy$merge %in% merge.freq.table.gt.1$Var1,] - #patient2.fuzzy = patient2.fuzzy[patient2.fuzzy$merge %in% merge.freq.table.gt.1$Var1,] - + patient.fuzzy = rbind(patient1.fuzzy, patient2.fuzzy) patient.fuzzy = patient.fuzzy[order(nchar(patient.fuzzy$Clone_Sequence)),] - + merge.list = list() - + merge.list[["second"]] = vector() - - link.count = 1 - + + link.count = 1 + while(nrow(patient.fuzzy) > 1){ first.merge = patient.fuzzy[1,"merge"] first.clone.sequence = patient.fuzzy[1,"Clone_Sequence"] first.sample = patient.fuzzy[1,"Sample"] merge.filter = first.merge == patient.fuzzy$merge - + #length.filter = nchar(patient.fuzzy$Clone_Sequence) - nchar(first.clone.sequence) <= 9 - + first.sample.filter = first.sample == patient.fuzzy$Sample second.sample.filter = first.sample != patient.fuzzy$Sample - + #first match same sample, sum to a single row, same for other sample #then merge rows like 'normal' - + sequence.filter = grepl(paste("^", first.clone.sequence, sep=""), patient.fuzzy$Clone_Sequence) - - - + + + #match.filter = merge.filter & grepl(first.clone.sequence, patient.fuzzy$Clone_Sequence) & length.filter & sample.filter first.match.filter = merge.filter & sequence.filter & first.sample.filter second.match.filter = merge.filter & sequence.filter & second.sample.filter - + first.rows = patient.fuzzy[first.match.filter,] second.rows = patient.fuzzy[second.match.filter,] - + first.rows.v = table(first.rows$V_Segment_Major_Gene) first.rows.v = names(first.rows.v[which.max(first.rows.v)]) first.rows.j = table(first.rows$J_Segment_Major_Gene) first.rows.j = names(first.rows.j[which.max(first.rows.j)]) - + first.sum = data.frame(merge = first.clone.sequence, Patient = patient, Receptor = first.rows[1,"Receptor"], @@ -249,61 +239,65 @@ normalized_read_count = sum(first.rows$normalized_read_count), paste = first.rows[1,"paste"], min_cell_paste = first.rows[1,"min_cell_paste"]) - + if(nrow(second.rows) > 0){ second.rows.v = table(second.rows$V_Segment_Major_Gene) second.rows.v = names(second.rows.v[which.max(second.rows.v)]) second.rows.j = table(second.rows$J_Segment_Major_Gene) second.rows.j = names(second.rows.j[which.max(second.rows.j)]) - + second.sum = data.frame(merge = first.clone.sequence, - Patient = patient, - Receptor = second.rows[1,"Receptor"], - Sample = second.rows[1,"Sample"], - Cell_Count = second.rows[1,"Cell_Count"], - Clone_Molecule_Count_From_Spikes = sum(second.rows$Clone_Molecule_Count_From_Spikes), - Log10_Frequency = log10(sum(second.rows$Frequency)), - Total_Read_Count = sum(second.rows$Total_Read_Count), - dsPerM = sum(second.rows$dsPerM), - J_Segment_Major_Gene = second.rows.j, - V_Segment_Major_Gene = second.rows.v, - Clone_Sequence = first.clone.sequence, - CDR3_Sense_Sequence = second.rows[1,"CDR3_Sense_Sequence"], - Related_to_leukemia_clone = F, - Frequency = sum(second.rows$Frequency), - locus_V = second.rows[1,"locus_V"], - locus_J = second.rows[1,"locus_J"], - min_cell_count = second.rows[1,"min_cell_count"], - normalized_read_count = sum(second.rows$normalized_read_count), - paste = second.rows[1,"paste"], - min_cell_paste = second.rows[1,"min_cell_paste"]) - + Patient = patient, + Receptor = second.rows[1,"Receptor"], + Sample = second.rows[1,"Sample"], + Cell_Count = second.rows[1,"Cell_Count"], + Clone_Molecule_Count_From_Spikes = sum(second.rows$Clone_Molecule_Count_From_Spikes), + Log10_Frequency = log10(sum(second.rows$Frequency)), + Total_Read_Count = sum(second.rows$Total_Read_Count), + dsPerM = sum(second.rows$dsPerM), + J_Segment_Major_Gene = second.rows.j, + V_Segment_Major_Gene = second.rows.v, + Clone_Sequence = first.clone.sequence, + CDR3_Sense_Sequence = second.rows[1,"CDR3_Sense_Sequence"], + Related_to_leukemia_clone = F, + Frequency = sum(second.rows$Frequency), + locus_V = second.rows[1,"locus_V"], + locus_J = second.rows[1,"locus_J"], + min_cell_count = second.rows[1,"min_cell_count"], + normalized_read_count = sum(second.rows$normalized_read_count), + paste = second.rows[1,"paste"], + min_cell_paste = second.rows[1,"min_cell_paste"]) + + #print(names(patientMerge)) + #print(merge(first.sum, second.sum, by="merge")) patientMerge = rbind(patientMerge, merge(first.sum, second.sum, by="merge")) + #print("test2") patient.fuzzy = patient.fuzzy[!(first.match.filter | second.match.filter),] - + hidden.clone.sequences = c(first.rows[-1,"Clone_Sequence"], second.rows[second.rows$Clone_Sequence != first.clone.sequence,"Clone_Sequence"]) merge.list[["second"]] = append(merge.list[["second"]], hidden.clone.sequences) - + tmp.rows = rbind(first.rows, second.rows) + #print("test3") tmp.rows = tmp.rows[order(nchar(tmp.rows$Clone_Sequence)),] #add to the scatterplot data scatterplot.row = first.sum[,scatterplot_data_columns] - scatterplot.row$type = paste(first.sum[,"Sample"], "In Both") - scatterplot.row$link = link.count - scatterplot.row$on = onShort - - scatterplot_data = rbind(scatterplot_data, scatterplot.row) + scatterplot.row$type = paste(first.sum[,"Sample"], "In Both") + scatterplot.row$link = link.count + scatterplot.row$on = onShort + + scatterplot_data = rbind(scatterplot_data, scatterplot.row) scatterplot.row = second.sum[,scatterplot_data_columns] - scatterplot.row$type = paste(second.sum[,"Sample"], "In Both") - scatterplot.row$link = link.count - scatterplot.row$on = onShort - - scatterplot_data = rbind(scatterplot_data, scatterplot.row) - - #write some information about the match to a log file + scatterplot.row$type = paste(second.sum[,"Sample"], "In Both") + scatterplot.row$link = link.count + scatterplot.row$on = onShort + + scatterplot_data = rbind(scatterplot_data, scatterplot.row) + + #write some information about the match to a log file if (nrow(first.rows) > 1 | nrow(second.rows) > 1) { cat(paste("<tr><td>", patient, " row ", 1:nrow(tmp.rows), "</td><td>", tmp.rows$Sample, ":</td><td>", tmp.rows$Clone_Sequence, "</td><td>", tmp.rows$normalized_read_count, "</td></tr>", sep=""), file="multiple_matches.html", append=T) } else { @@ -314,7 +308,7 @@ #cat(paste("<tr><td>", patient, " row ", 1:nrow(tmp.rows), "</td><td>", tmp.rows$Sample, ":</td><td>", tmp.rows$Clone_Sequence, "</td><td>", tmp.rows$normalized_read_count, "</td></tr>", sep=""), file="single_matches.html", append=T) } } - + } else if(nrow(first.rows) > 1) { if(patient1[1,"Sample"] == first.sample){ patient1 = patient1[!(patient1$Clone_Sequence %in% first.rows$Clone_Sequence),] @@ -323,31 +317,31 @@ patient2 = patient2[!(patient2$Clone_Sequence %in% first.rows$Clone_Sequence),] patient2 = rbind(patient2, first.sum) } - + hidden.clone.sequences = c(first.rows[-1,"Clone_Sequence"]) merge.list[["second"]] = append(merge.list[["second"]], hidden.clone.sequences) - + patient.fuzzy = patient.fuzzy[-first.match.filter,] #add to the scatterplot data scatterplot.row = first.sum[,scatterplot_data_columns] - scatterplot.row$type = first.sum[,"Sample"] - scatterplot.row$link = link.count - scatterplot.row$on = onShort - - scatterplot_data = rbind(scatterplot_data, scatterplot.row) - + scatterplot.row$type = first.sum[,"Sample"] + scatterplot.row$link = link.count + scatterplot.row$on = onShort + + scatterplot_data = rbind(scatterplot_data, scatterplot.row) + cat(paste("<tr bgcolor='#DDF'><td>", patient, " row ", 1:nrow(first.rows), "</td><td>", first.rows$Sample, ":</td><td>", first.rows$Clone_Sequence, "</td><td>", first.rows$normalized_read_count, "</td></tr>", sep=""), file="single_matches.html", append=T) } else { patient.fuzzy = patient.fuzzy[-1,] #add to the scatterplot data scatterplot.row = first.sum[,scatterplot_data_columns] - scatterplot.row$type = first.sum[,"Sample"] - scatterplot.row$link = link.count - scatterplot.row$on = onShort - - scatterplot_data = rbind(scatterplot_data, scatterplot.row) + scatterplot.row$type = first.sum[,"Sample"] + scatterplot.row$link = link.count + scatterplot.row$on = onShort + + scatterplot_data = rbind(scatterplot_data, scatterplot.row) } link.count = link.count + 1 } @@ -360,10 +354,10 @@ scatter_locus_data_list[[patient]] <<- scatterplot_data cat(paste("<td>", nrow(patient1), " in ", oneSample, " and ", nrow(patient2), " in ", twoSample, ", ", nrow(patientMerge), " in both (finding both took ", (proc.time() - start.time)[[3]], "s)</td></tr>", sep=""), file=logfile, append=T) } - + patient1 = patient1[!(patient1$Clone_Sequence %in% patient.merge.list.second[[patient]]),] patient2 = patient2[!(patient2$Clone_Sequence %in% patient.merge.list.second[[patient]]),] - + patientMerge$thresholdValue = pmax(patientMerge[,onx], patientMerge[,ony]) #patientMerge$thresholdValue = pmin(patientMerge[,onx], patientMerge[,ony]) @@ -377,9 +371,9 @@ #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="") + threshhold = product[iter,"interval"] + V_Segment = paste(".*", as.character(product[iter,"V_Segments"]), ".*", sep="") + J_Segment = paste(".*", as.character(product[iter,"J_Segments"]), ".*", 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$thresholdValue > threshhold) #highest of both is higher than threshold one = (grepl(V_Segment, patient1$V_Segment_Major_Gene) & grepl(J_Segment, patient1$J_Segment_Major_Gene) & patient1[,on] > threshhold & !(patient1$merge %in% patientMerge[both,]$merge)) @@ -392,50 +386,50 @@ locussum1 = append(locussum1, sum(patient1[(grepl(V_Segment, patient1$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)),]$normalized_read_count)) #threshhold = 0 - if(threshhold != 0){ + if(threshhold != 0 | T){ if(sum(one) > 0){ dfOne = patient1[one,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfOne) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone Sequence", "Related_to_leukemia_clone") - filenameOne = paste(oneSample, "_", product[iter, titleIndex], "_", threshhold, sep="") + filenameOne = paste(oneSample, "_", product[iter, "Titles"], "_", threshhold, sep="") write.table(dfOne, file=paste(filenameOne, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } if(sum(two) > 0){ dfTwo = patient2[two,c("V_Segment_Major_Gene", "J_Segment_Major_Gene", "normalized_read_count", "Frequency", "Clone_Sequence", "Related_to_leukemia_clone")] colnames(dfTwo) = c("Proximal segment", "Distal segment", "normalized_read_count", "Frequency", "Clone Sequence", "Related_to_leukemia_clone") - filenameTwo = paste(twoSample, "_", product[iter, titleIndex], "_", threshhold, sep="") + filenameTwo = paste(twoSample, "_", product[iter, "Titles"], "_", threshhold, sep="") write.table(dfTwo, file=paste(filenameTwo, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } } else { scatterplot_locus_data = scatterplot_data[grepl(V_Segment, scatterplot_data$V_Segment_Major_Gene) & grepl(J_Segment, scatterplot_data$J_Segment_Major_Gene),] if(nrow(scatterplot_locus_data) > 0){ - scatterplot_locus_data$Rearrangement = product[iter, titleIndex] + scatterplot_locus_data$Rearrangement = product[iter, "Titles"] } - + p = NULL - print(paste("nrow scatterplot_locus_data", nrow(scatterplot_locus_data))) + #print(paste("nrow scatterplot_locus_data", nrow(scatterplot_locus_data))) if(nrow(scatterplot_locus_data) != 0){ if(on == "normalized_read_count"){ - write.table(scatterplot_locus_data, file=paste(oneSample, twoSample, product[iter, titleIndex], "scatterplot_locus_data.txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) + write.table(scatterplot_locus_data, file=paste(oneSample, twoSample, product[iter, "Titles"], "scatterplot_locus_data.txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) scales = 10^(0:6) #(0:ceiling(log10(max(scatterplot_locus_data$normalized_read_count)))) p = ggplot(scatterplot_locus_data, aes(factor(reorder(type, type.order)), normalized_read_count, group=link)) + geom_line() + scale_y_log10(breaks=scales,labels=scales, limits=c(1,1e6)) + scale_x_discrete(breaks=levels(scatterplot_data$type), labels=levels(scatterplot_data$type), drop=FALSE) } else { p = ggplot(scatterplot_locus_data, aes(factor(reorder(type, type.order)), Frequency, group=link)) + geom_line() + scale_y_log10(limits=c(0.0001,100), breaks=c(0.0001, 0.001, 0.01, 0.1, 1, 10, 100), labels=c("0.0001", "0.001", "0.01", "0.1", "1", "10", "100")) + scale_x_discrete(breaks=levels(scatterplot_data$type), labels=levels(scatterplot_data$type), drop=FALSE) } p = p + geom_point(aes(colour=type), position="dodge") - p = p + xlab("In one or both samples") + ylab(onShort) + ggtitle(paste(patient1[1,patientIndex], patient1[1,sampleIndex], patient2[1,sampleIndex], onShort, product[iter, titleIndex])) + p = p + xlab("In one or both samples") + ylab(onShort) + ggtitle(paste(patient1[1,"Patient"], patient1[1,"Sample"], patient2[1,"Sample"], onShort, product[iter, "Titles"])) } else { - p = ggplot(NULL, aes(x=c("In one", "In Both"),y=0)) + geom_blank(NULL) + xlab("In one or both of the samples") + ylab(onShort) + ggtitle(paste(patient1[1,patientIndex], patient1[1,sampleIndex], patient2[1,sampleIndex], onShort, product[iter, titleIndex])) + p = ggplot(NULL, aes(x=c("In one", "In Both"),y=0)) + geom_blank(NULL) + xlab("In one or both of the samples") + ylab(onShort) + ggtitle(paste(patient1[1,"Patient"], patient1[1,"Sample"], patient2[1,"Sample"], onShort, product[iter, titleIndex])) } - png(paste(patient1[1,patientIndex], "_", patient1[1,sampleIndex], "_", patient2[1,sampleIndex], "_", onShort, "_", product[iter, titleIndex],"_scatter.png", sep="")) + png(paste(patient1[1,"Patient"], "_", patient1[1,"Sample"], "_", patient2[1,"Sample"], "_", onShort, "_", product[iter, "Titles"],"_scatter.png", sep="")) print(p) dev.off() } 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="") + filenameBoth = paste(oneSample, "_", twoSample, "_", product[iter, "Titles"], "_", threshhold, sep="") write.table(dfBoth, file=paste(filenameBoth, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) } } @@ -487,20 +481,27 @@ print(plt) dev.off() } + if(length(patients) > 0){ 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))) - lapply(patients, FUN=patientCountOnColumn, product = product, interval=interval, on="Frequency", appendtxt=T) + for (current_patient in patients){ + print(paste("Started working", unique(current_patient$Patient), "Frequency analysis")) + patientCountOnColumn(current_patient, product=product, interval=interval, on="Frequency", appendtxt=T) + } cat("<tr><td>Starting Cell Count analysis</td></tr>", file=logfile, append=T) interval = intervalReads intervalOrder = data.frame("interval"=paste(">", interval, sep=""), "intervalOrder"=1:length(interval)) product = data.frame("Titles"=rep(Titles, each=length(interval)), "interval"=rep(interval, times=10), "V_Segments"=rep(V_Segments, each=length(interval)), "J_Segments"=rep(J_Segments, each=length(interval))) - lapply(patients, FUN=patientCountOnColumn, product = product, interval=interval, on="normalized_read_count") + for (current_patient in patients){ + print(paste("Started working on ", unique(current_patient$Patient), "Read Count analysis")) + patientCountOnColumn(current_patient, product=product, interval=interval, on="normalized_read_count") + } } if(nrow(single_patients) > 0){ scales = 10^(0:6) #(0:ceiling(log10(max(scatterplot_locus_data$normalized_read_count)))) @@ -936,18 +937,20 @@ } p = NULL if(nrow(scatterplot_locus_data) != 0){ - if(on == "normalized_read_count"){ - scales = 10^(0:6) #(0:ceiling(log10(max(scatterplot_locus_data$normalized_read_count)))) - p = ggplot(scatterplot_locus_data, aes(type, normalized_read_count)) + scale_y_log10(breaks=scales,labels=scales, limits=c(1, 1e6)) + filename.scatter = paste(label1, "_", label2, "_", label3, "_", product[iter, titleIndex], "_scatter_", threshhold, sep="") + write.table(scatterplot_locus_data, file=paste(filename.scatter, ".txt", sep=""), quote=F, sep="\t", dec=",", row.names=F, col.names=T) + if(on == "normalized_read_count"){ + scales = 10^(0:6) #(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, limits=c(1, 1e6)) + } else { + p = ggplot(scatterplot_locus_data, aes(type, Frequency)) + scale_y_log10(limits=c(0.0001,100), breaks=c(0.0001, 0.001, 0.01, 0.1, 1, 10, 100), labels=c("0.0001", "0.001", "0.01", "0.1", "1", "10", "100")) + expand_limits(y=c(0,100)) + #p = ggplot(scatterplot_locus_data, aes(type, Frequency)) + scale_y_continuous(limits = c(0, 100)) + expand_limits(y=c(0,100)) + } + 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(scatterplot_locus_data, aes(type, Frequency)) + scale_y_log10(limits=c(0.0001,100), breaks=c(0.0001, 0.001, 0.01, 0.1, 1, 10, 100), labels=c("0.0001", "0.001", "0.01", "0.1", "1", "10", "100")) + expand_limits(y=c(0,100)) - #p = ggplot(scatterplot_locus_data, aes(type, Frequency)) + scale_y_continuous(limits = c(0, 100)) + expand_limits(y=c(0,100)) + 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])) } - 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()