annotate shm_csr.r @ 0:c33d93683a09 draft

Uploaded

1 library(data.table)
2 library(ggplot2)
3 library(reshape2)
4 
5 args <- commandArgs(trailingOnly = TRUE)
6 
7 input = args[1]
8 genes = unlist(strsplit(args[2], ","))
9 outputdir = args[3]
10 include_fr1 = ifelse(args[4] == "yes", T, F)
11 setwd(outputdir)
12 
13 dat = read.table(input, header=T, sep="\t", fill=T, stringsAsFactors=F)
14 
15 if(length(dat$Sequence.ID) == 0){
16   setwd(outputdir)
17   result = data.frame(x = rep(0, 5), y = rep(0, 5), z = rep(NA, 5))
18   row.names(result) = c("Number of Mutations (%)", "Transition (%)", "Transversions (%)", "Transitions at G C (%)", "Targeting of C G (%)")
19   write.table(x=result, file="mutations.txt", sep=",",quote=F,row.names=T,col.names=F)
20   transitionTable = data.frame(A=rep(0, 4),C=rep(0, 4),G=rep(0, 4),T=rep(0, 4))
21   row.names(transitionTable) = c("A", "C", "G", "T")
22   transitionTable["A","A"] = NA
23   transitionTable["C","C"] = NA
24   transitionTable["G","G"] = NA
25   transitionTable["T","T"] = NA
26   write.table(x=transitionTable, file="transitions.txt", sep=",",quote=F,row.names=T,col.names=NA)
27   cat("0", file="n.txt")
28   stop("No data")
29 }
30 
31 cleanup_columns = c("FR1.IMGT.c.a",
32                     "FR2.IMGT.g.t",
33                     "CDR1.IMGT.Nb.of.nucleotides",
34                     "CDR2.IMGT.t.a",
35                     "FR1.IMGT.c.g",
36                     "CDR1.IMGT.c.t",
37                     "FR2.IMGT.a.c",
38                     "FR2.IMGT.Nb.of.mutations",
39                     "FR2.IMGT.g.c",
40                     "FR2.IMGT.a.g",
41                     "FR3.IMGT.t.a",
42                     "FR3.IMGT.t.c",
43                     "FR2.IMGT.g.a",
44                     "FR3.IMGT.c.g",
45                     "FR1.IMGT.Nb.of.mutations",
46                     "CDR1.IMGT.g.a",
47                     "CDR1.IMGT.t.g",
48                     "CDR1.IMGT.g.c",
49                     "CDR2.IMGT.Nb.of.nucleotides",
50                     "FR2.IMGT.a.t",
51                     "CDR1.IMGT.Nb.of.mutations",
52                     "CDR3.IMGT.Nb.of.nucleotides",
53                     "CDR1.IMGT.a.g",
54                     "FR3.IMGT.a.c",
55                     "FR1.IMGT.g.a",
56                     "FR3.IMGT.a.g",
57                     "FR1.IMGT.a.t",
58                     "CDR2.IMGT.a.g",
59                     "CDR2.IMGT.Nb.of.mutations",
60                     "CDR2.IMGT.g.t",
61                     "CDR2.IMGT.a.c",
62                     "CDR1.IMGT.t.c",
63                     "FR3.IMGT.g.c",
64                     "FR1.IMGT.g.t",
65                     "FR3.IMGT.g.t",
66                     "CDR1.IMGT.a.t",
67                     "FR1.IMGT.a.g",
68                     "FR3.IMGT.a.t",
69                     "FR3.IMGT.Nb.of.nucleotides",
70                     "FR2.IMGT.t.c",
71                     "CDR2.IMGT.g.a",
72                     "FR2.IMGT.t.a",
73                     "CDR1.IMGT.t.a",
74                     "FR2.IMGT.t.g",
75                     "FR3.IMGT.t.g",
76                     "FR2.IMGT.Nb.of.nucleotides",
77                     "FR1.IMGT.t.a",
78                     "FR1.IMGT.t.g",
79                     "FR3.IMGT.c.t",
80                     "FR1.IMGT.t.c",
81                     "CDR2.IMGT.a.t",
82                     "FR2.IMGT.c.t",
83                     "CDR1.IMGT.g.t",
84                     "CDR2.IMGT.t.g",
85                     "FR1.IMGT.Nb.of.nucleotides",
86                     "CDR1.IMGT.c.g",
87                     "CDR2.IMGT.t.c",
88                     "FR3.IMGT.g.a",
89                     "CDR1.IMGT.a.c",
90                     "FR2.IMGT.c.a",
91                     "FR3.IMGT.Nb.of.mutations",
92                     "FR2.IMGT.c.g",
93                     "CDR2.IMGT.g.c",
94                     "FR1.IMGT.g.c",
95                     "CDR2.IMGT.c.t",
96                     "FR3.IMGT.c.a",
97                     "CDR1.IMGT.c.a",
98                     "CDR2.IMGT.c.g",
99                     "CDR2.IMGT.c.a",
00                     "FR1.IMGT.c.t",
01                     "FR1.IMGT.Nb.of.silent.mutations",
02                     "FR2.IMGT.Nb.of.silent.mutations",
03                     "FR3.IMGT.Nb.of.silent.mutations",
04                     "FR1.IMGT.Nb.of.nonsilent.mutations",
05                     "FR2.IMGT.Nb.of.nonsilent.mutations",
06                     "FR3.IMGT.Nb.of.nonsilent.mutations")
07 
08 
09 print("Cleaning up columns")
10 for(col in cleanup_columns){
11   dat[,col] = gsub("\\(.*\\)", "", dat[,col])
12   #dat[dat[,col] == "",] = "0"
13   dat[,col] = as.numeric(dat[,col])
14   dat[is.na(dat[,col]),col] = 0
15 }
16 
17 regions = c("FR1", "CDR1", "FR2", "CDR2", "FR3")
18 if(!include_fr1){
19 	regions = c("CDR1", "FR2", "CDR2", "FR3")
20 }
21 
22 sum_by_row = function(x, columns) { sum(as.numeric(x[columns]), na.rm=T) }
23 
24 print("aggregating data into new columns")
25 
26 VRegionMutations_columns = paste(regions, ".IMGT.Nb.of.mutations", sep="")
27 dat$VRegionMutations =  apply(dat, FUN=sum_by_row, 1, columns=VRegionMutations_columns)
28 
29 VRegionNucleotides_columns = paste(regions, ".IMGT.Nb.of.nucleotides", sep="")
30 dat$FR3.IMGT.Nb.of.nucleotides = nchar(dat$FR3.IMGT.seq)
31 dat$VRegionNucleotides =  apply(dat, FUN=sum_by_row, 1, columns=VRegionNucleotides_columns)
32 
33 transitionMutations_columns = paste(rep(regions, each=4), c(".IMGT.a.g", ".IMGT.g.a", ".IMGT.c.t", ".IMGT.t.c"), sep="")
34 dat$transitionMutations = apply(dat, FUN=sum_by_row, 1, columns=transitionMutations_columns)
35 
36 transversionMutations_columns = paste(rep(regions, each=8), c(".IMGT.a.c",".IMGT.c.a",".IMGT.a.t",".IMGT.t.a",".IMGT.g.c",".IMGT.c.g",".IMGT.g.t",".IMGT.t.g"), sep="")
37 dat$transversionMutations = apply(dat, FUN=sum_by_row, 1, columns=transversionMutations_columns)
38 
39 
40 transitionMutationsAtGC_columns = paste(rep(regions, each=2), c(".IMGT.g.a",".IMGT.c.t"), sep="")
41 dat$transitionMutationsAtGC = apply(dat, FUN=sum_by_row, 1, columns=transitionMutationsAtGC_columns)
42 
43 
44 totalMutationsAtGC_columns = paste(rep(regions, each=6), c(".IMGT.c.g",".IMGT.c.t",".IMGT.c.a",".IMGT.g.c",".IMGT.g.a",".IMGT.g.t"), sep="")
45 #totalMutationsAtGC_columns = paste(rep(regions, each=6), c(".IMGT.g.a",".IMGT.c.t",".IMGT.c.a",".IMGT.c.g",".IMGT.g.t"), sep="")
46 dat$totalMutationsAtGC = apply(dat, FUN=sum_by_row, 1, columns=totalMutationsAtGC_columns)
47 
48 transitionMutationsAtAT_columns = paste(rep(regions, each=2), c(".IMGT.a.g",".IMGT.t.c"), sep="")
49 dat$transitionMutationsAtAT = apply(dat, FUN=sum_by_row, 1, columns=transitionMutationsAtAT_columns)
50 
51 totalMutationsAtAT_columns = paste(rep(regions, each=6), c(".IMGT.a.g",".IMGT.a.c",".IMGT.a.t",".IMGT.t.g",".IMGT.t.c",".IMGT.t.a"), sep="")
52 #totalMutationsAtAT_columns = paste(rep(regions, each=5), c(".IMGT.a.g",".IMGT.t.c",".IMGT.a.c",".IMGT.g.c",".IMGT.t.g"), sep="")
53 dat$totalMutationsAtAT = apply(dat, FUN=sum_by_row, 1, columns=totalMutationsAtAT_columns)
54 
55 
56 FRRegions = regions[grepl("FR", regions)]
57 CDRRegions = regions[grepl("CDR", regions)]
58 
59 FR_silentMutations_columns = paste(FRRegions, ".IMGT.Nb.of.silent.mutations", sep="")
60 dat$silentMutationsFR = apply(dat, FUN=sum_by_row, 1, columns=FR_silentMutations_columns)
61 
62 CDR_silentMutations_columns = paste(CDRRegions, ".IMGT.Nb.of.silent.mutations", sep="")
63 dat$silentMutationsCDR = apply(dat, FUN=sum_by_row, 1, columns=CDR_silentMutations_columns)
64 
65 FR_nonSilentMutations_columns = paste(FRRegions, ".IMGT.Nb.of.nonsilent.mutations", sep="")
66 dat$nonSilentMutationsFR = apply(dat, FUN=sum_by_row, 1, columns=FR_nonSilentMutations_columns)
67 
68 CDR_nonSilentMutations_columns = paste(CDRRegions, ".IMGT.Nb.of.nonsilent.mutations", sep="")
69 dat$nonSilentMutationsCDR = apply(dat, FUN=sum_by_row, 1, columns=CDR_nonSilentMutations_columns)
70 
71 mutation.sum.columns = c("Sequence.ID", "VRegionMutations", "VRegionNucleotides", "transitionMutations", "transversionMutations", "transitionMutationsAtGC", "transitionMutationsAtAT", "silentMutationsFR", "nonSilentMutationsFR", "silentMutationsCDR", "nonSilentMutationsCDR")
72 
73 write.table(dat[,mutation.sum.columns], "mutation_by_id.txt", sep="\t",quote=F,row.names=F,col.names=T)
74 
75 setwd(outputdir)
76 
77 base.order = data.frame(base=c("A", "T", "C", "G"), order=1:4)
78 
79 calculate_result = function(i, gene, dat, matrx, f, fname, name){
80 	tmp = dat[grepl(paste("^", gene, ".*", sep=""), dat$best_match),]
81 
82 	j = i - 1
83 	x = (j * 3) + 1
84 	y = (j * 3) + 2
85 	z = (j * 3) + 3
86 	 
87 	if(nrow(tmp) > 0){
88 	  
89 		if(fname == "sum"){
90 		matrx[1,x] = round(f(tmp$VRegionMutations, na.rm=T), digits=1)
91 		matrx[1,y] = round(f(tmp$VRegionNucleotides, na.rm=T), digits=1)
92 		matrx[1,z] = round(f(matrx[1,x] / matrx[1,y]) * 100, digits=1)
93 		} else {
94 		matrx[1,x] = round(f(tmp$VRegionMutations, na.rm=T), digits=1)
95 		matrx[1,y] = round(f(tmp$VRegionNucleotides, na.rm=T), digits=1)
96 		matrx[1,z] = round(f(tmp$VRegionMutations / tmp$VRegionNucleotides) * 100, digits=1)
97 		}
98 
99 		matrx[2,x] = round(f(tmp$transitionMutations, na.rm=T), digits=1)
00 		matrx[2,y] = round(f(tmp$VRegionMutations, na.rm=T), digits=1)
01 		matrx[2,z] = round(matrx[2,x] / matrx[2,y] * 100, digits=1)
02 
03 		matrx[3,x] = round(f(tmp$transversionMutations, na.rm=T), digits=1)
04 		matrx[3,y] = round(f(tmp$VRegionMutations, na.rm=T), digits=1)
05 		matrx[3,z] = round(matrx[3,x] / matrx[3,y] * 100, digits=1)
06 
07 		matrx[4,x] = round(f(tmp$transitionMutationsAtGC, na.rm=T), digits=1)
08 		matrx[4,y] = round(f(tmp$totalMutationsAtGC, na.rm=T), digits=1)
09 		matrx[4,z] = round(matrx[4,x] / matrx[4,y] * 100, digits=1)
10 
11 		matrx[5,x] = round(f(tmp$totalMutationsAtGC, na.rm=T), digits=1)
12 		matrx[5,y] = round(f(tmp$VRegionMutations, na.rm=T), digits=1)
13 		matrx[5,z] = round(matrx[5,x] / matrx[5,y] * 100, digits=1)
14 
15 		matrx[6,x] = round(f(tmp$transitionMutationsAtAT, na.rm=T), digits=1)
16 		matrx[6,y] = round(f(tmp$totalMutationsAtAT, na.rm=T), digits=1)
17 		matrx[6,z] = round(matrx[6,x] / matrx[6,y] * 100, digits=1)
18 
19 		matrx[7,x] = round(f(tmp$totalMutationsAtAT, na.rm=T), digits=1)
20 		matrx[7,y] = round(f(tmp$VRegionMutations, na.rm=T), digits=1)
21 		matrx[7,z] = round(matrx[7,x] / matrx[7,y] * 100, digits=1)
22 
23 		matrx[8,x] = round(f(tmp$nonSilentMutationsFR, na.rm=T), digits=1)
24 		matrx[8,y] = round(f(tmp$silentMutationsFR, na.rm=T), digits=1)
25 		matrx[8,z] = round(matrx[8,x] / matrx[8,y], digits=1)
26 
27 		matrx[9,x] = round(f(tmp$nonSilentMutationsCDR, na.rm=T), digits=1)
28 		matrx[9,y] = round(f(tmp$silentMutationsCDR, na.rm=T), digits=1)
29 		matrx[9,z] = round(matrx[9,x] / matrx[9,y], digits=1)
30 
31 		if(fname == "sum"){
32 			matrx[10,x] = round(f(rowSums(tmp[,c("FR2.IMGT.Nb.of.nucleotides", "FR3.IMGT.Nb.of.nucleotides")], na.rm=T)), digits=1)
33 			matrx[10,y] = round(f(tmp$VRegionNucleotides, na.rm=T), digits=1)
34 			matrx[10,z] = round(matrx[10,x] / matrx[10,y] * 100, digits=1)
35 
36 			matrx[11,x] = round(f(rowSums(tmp[,c("CDR1.IMGT.Nb.of.nucleotides", "CDR2.IMGT.Nb.of.nucleotides")], na.rm=T)), digits=1)
37 			matrx[11,y] = round(f(tmp$VRegionNucleotides, na.rm=T), digits=1)
38 			matrx[11,z] = round(matrx[11,x] / matrx[11,y] * 100, digits=1)
39 		}
40 	}
41   
42 	transitionTable = data.frame(A=zeros,C=zeros,G=zeros,T=zeros)
43 	row.names(transitionTable) = c("A", "C", "G", "T")
44 	transitionTable["A","A"] = NA
45 	transitionTable["C","C"] = NA
46 	transitionTable["G","G"] = NA
47 	transitionTable["T","T"] = NA
48 
49 	if(nrow(tmp) > 0){
50 		for(nt1 in nts){
51 			for(nt2 in nts){
52 				if(nt1 == nt2){
53 					next
54 				}
55 				NT1 = LETTERS[letters == nt1]
56 				NT2 = LETTERS[letters == nt2]
57 				FR1 = paste("FR1.IMGT.", nt1, ".", nt2, sep="")
58 				CDR1 = paste("CDR1.IMGT.", nt1, ".", nt2, sep="")
59 				FR2 = paste("FR2.IMGT.", nt1, ".", nt2, sep="")
60 				CDR2 = paste("CDR2.IMGT.", nt1, ".", nt2, sep="")
61 				FR3 = paste("FR3.IMGT.", nt1, ".", nt2, sep="")
62 				if(include_fr1){
63 					transitionTable[NT1,NT2] = sum(tmp[,c(FR1, CDR1, FR2, CDR2, FR3)])
64 				} else {
65 					transitionTable[NT1,NT2] = sum(tmp[,c(CDR1, FR2, CDR2, FR3)])
66 				}
67 			}
68 		}
69 		transition = transitionTable
70 		transition$id = names(transition)
71 		
72 		transition2 = melt(transition, id.vars="id")
73 		
74 		transition2 = merge(transition2, base.order, by.x="id", by.y="base")
75 		transition2 = merge(transition2, base.order, by.x="variable", by.y="base")
76 
77 		transition2[is.na(transition2$value),]$value = 0
78 		
79 		if(!all(transition2$value == 0)){ #having rows of data but a transition table filled with 0 is bad
80 
81 			print("Plotting stacked transition")
82 
83 			png(filename=paste("transitions_stacked_", name, ".png", sep=""))
84 			p = ggplot(transition2, aes(factor(reorder(id, order.x)), y=value, fill=factor(reorder(variable, order.y)))) + geom_bar(position="fill", stat="identity", colour="black") #stacked bar
85 			p = p + xlab("From base") + ylab("To base") + ggtitle("Mutations frequency from base to base") + guides(fill=guide_legend(title=NULL))
86 			p = p + theme(panel.background = element_rect(fill = "white", colour="black")) + scale_fill_manual(values=c("A" = "blue4", "G" = "lightblue1", "C" = "olivedrab3", "T" = "olivedrab4"))
87 			#p = p + scale_colour_manual(values=c("A" = "black", "G" = "black", "C" = "black", "T" = "black"))
88 			print(p)
89 			dev.off()
90 
91 			print("Plotting heatmap transition")
92 
93 			png(filename=paste("transitions_heatmap_", name, ".png", sep=""))
94 			p = ggplot(transition2, aes(factor(reorder(id, order.x)), factor(reorder(variable, order.y)))) + geom_tile(aes(fill = value)) + scale_fill_gradient(low="white", high="steelblue") #heatmap
95 			p = p + xlab("From base") + ylab("To base") + ggtitle("Mutations frequency from base to base")  + theme(panel.background = element_rect(fill = "white", colour="black"))
96 			print(p)
97 			dev.off()
98 		} else {
99 			print("No data to plot")
00 		}
01 	}
02 
03 	#print(paste("writing value file: ", name, "_", fname, "_value.txt" ,sep=""))
04 
05 	write.table(x=transitionTable, file=paste("transitions_", name ,"_", fname, ".txt", sep=""), sep=",",quote=F,row.names=T,col.names=NA)
06 	write.table(x=tmp[,c("Sequence.ID", "best_match", "chunk_hit_percentage", "nt_hit_percentage", "start_locations")], file=paste("matched_", name , "_", fname, ".txt", sep=""), sep="\t",quote=F,row.names=F,col.names=T)
07 
08 	cat(matrx[1,x], file=paste(name, "_", fname, "_value.txt" ,sep=""))
09 	cat(nrow(tmp), file=paste(name, "_", fname, "_n.txt" ,sep=""))
10 
11 	#print(paste(fname, name, nrow(tmp)))
12 
13 	matrx
14 }
15 
16 nts = c("a", "c", "g", "t")
17 zeros=rep(0, 4)
18 
19 funcs = c(median, sum, mean)
20 fnames = c("median", "sum", "mean")
21 
22 print("Creating result tables")
23 
24 for(i in 1:length(funcs)){
25 	func = funcs[[i]]
26 	fname = fnames[[i]]
27 	
28 	rows = 9
29 	if(fname == "sum"){
30 		rows = 11
31 	}
32 	matrx = matrix(data = 0, ncol=((length(genes) + 1) * 3),nrow=rows)
33 	
34 	for(i in 1:length(genes)){
35 		print(paste("Creating table for", fname, genes[i]))
36 		matrx = calculate_result(i, genes[i], dat, matrx, func, fname, genes[i])
37 	}
38 
39 	matrx = calculate_result(i + 1, ".*", dat[!grepl("unmatched", dat$best_match),], matrx, func, fname, name="all")
40 	
41 	result = data.frame(matrx)
42 	if(fname == "sum"){
43 		row.names(result) = c("Number of Mutations (%)", "Transitions (%)", "Transversions (%)", "Transitions at G C (%)", "Targeting of C G (%)", "Transitions at A T (%)", "Targeting of A T (%)", "FR R/S (ratio)", "CDR R/S (ratio)", "nt in FR", "nt in CDR")
44 	} else {
45 		row.names(result) = c("Number of Mutations (%)", "Transitions (%)", "Transversions (%)", "Transitions at G C (%)", "Targeting of C G (%)", "Transitions at A T (%)", "Targeting of A T (%)", "FR R/S (ratio)", "CDR R/S (ratio)")
46 	}
47 
48 	write.table(x=result, file=paste("mutations_", fname, ".txt", sep=""), sep=",",quote=F,row.names=T,col.names=F)
49 }
50 
51 print("Adding median number of mutations to sum table")
52 
53 sum.table = read.table("mutations_sum.txt", sep=",", header=F)
54 median.table = read.table("mutations_median.txt", sep=",", header=F)
55 
56 new.table = sum.table[1,]
57 new.table[2,] = median.table[1,]
58 new.table[3:12,] = sum.table[2:11,]
59 new.table[,1] = as.character(new.table[,1])
60 new.table[2,1] = "Median of Number of Mutations (%)"
61 
62 #sum.table = sum.table[c("Number of Mutations (%)", "Median of Number of Mutations (%)", "Transition (%)", "Transversions (%)", "Transitions at G C (%)", "Targeting of C G (%)", "Transitions at A T (%)", "Targeting of A T (%)", "FR R/S (ratio)", "CDR R/S (ratio)", "nt in FR", "nt in CDR"),]
63 
64 write.table(x=new.table, file="mutations_sum.txt", sep=",",quote=F,row.names=F,col.names=F)
65 
66 
67 print("Plotting IGA piechart")
68 
69 dat = dat[!grepl("^unmatched", dat$best_match),]
70 
71 #blegh
72 genesForPlot = dat[grepl("IGA", dat$best_match),]$best_match
73 if(length(genesForPlot) > 0){
74 	genesForPlot = data.frame(table(genesForPlot))
75 	colnames(genesForPlot) = c("Gene","Freq")
76 	genesForPlot$label = paste(genesForPlot$Gene, "-", genesForPlot$Freq)
77 
78 	pc = ggplot(genesForPlot, aes(x = factor(1), y=Freq, fill=Gene))
79 	pc = pc + geom_bar(width = 1, stat = "identity") + scale_fill_manual(labels=genesForPlot$label, values=c("IGA1" = "lightblue1", "IGA2" = "blue4"))
80 	pc = pc + coord_polar(theta="y")
81 	pc = pc + theme(panel.background = element_rect(fill = "white", colour="black"))
82 	pc = pc + xlab(" ") + ylab(" ") + ggtitle(paste("IGA subclasses", "( n =", sum(genesForPlot$Freq), ")"))
83 	write.table(genesForPlot, "IGA.txt", sep="\t",quote=F,row.names=F,col.names=T)
84 
85 	png(filename="IGA.png")
86 	print(pc)
87 	dev.off()
88 }
89 
90 print("Plotting IGG piechart")
91 
92 genesForPlot = dat[grepl("IGG", dat$best_match),]$best_match
93 if(length(genesForPlot) > 0){
94 	genesForPlot = data.frame(table(genesForPlot))
95 	colnames(genesForPlot) = c("Gene","Freq")
96 	genesForPlot$label = paste(genesForPlot$Gene, "-", genesForPlot$Freq)
97 
98 	pc = ggplot(genesForPlot, aes(x = factor(1), y=Freq, fill=Gene))
99 	pc = pc + geom_bar(width = 1, stat = "identity") + scale_fill_manual(labels=genesForPlot$label, values=c("IGG1" = "olivedrab3", "IGG2" = "red", "IGG3" = "gold", "IGG4" = "darkred"))
00 	pc = pc + coord_polar(theta="y") 
01 	pc = pc + theme(panel.background = element_rect(fill = "white", colour="black"))
02 	pc = pc + xlab(" ") + ylab(" ") + ggtitle(paste("IGG subclasses", "( n =", sum(genesForPlot$Freq), ")"))
03 	write.table(genesForPlot, "IGG.txt", sep="\t",quote=F,row.names=F,col.names=T)
04 
05 	png(filename="IGG.png")
06 	print(pc)
07 	dev.off()
08 }
09 
10 
11 print("Plotting scatterplot")
12 
13 dat$percentage_mutations = round(dat$VRegionMutations / dat$VRegionNucleotides * 100, 2)
14 
15 p = ggplot(dat, aes(best_match, percentage_mutations))
16 p = p + geom_point(aes(colour=best_match), position="jitter") + geom_boxplot(aes(middle=mean(percentage_mutations)), alpha=0.1, outlier.shape = NA)
17 p = p + xlab("Subclass") + ylab("Frequency") + ggtitle("Frequency scatter plot") + theme(panel.background = element_rect(fill = "white", colour="black"))
18 p = p + scale_fill_manual(values=c("IGA1" = "lightblue1", "IGA2" = "blue4", "IGG1" = "olivedrab3", "IGG2" = "red", "IGG3" = "gold", "IGG4" = "darkred", "IGM" = "black"))
19 p = p + scale_colour_manual(values=c("IGA1" = "lightblue1", "IGA2" = "blue4", "IGG1" = "olivedrab3", "IGG2" = "red", "IGG3" = "gold", "IGG4" = "darkred", "IGM" = "black"))
20 
21 png(filename="scatter.png")
22 print(p)
23 dev.off()
24 
25 write.table(dat[,c("Sequence.ID", "best_match", "VRegionMutations", "VRegionNucleotides", "percentage_mutations")], "scatter.txt", sep="\t",quote=F,row.names=F,col.names=T)
26 
27 write.table(dat, input, sep="\t",quote=F,row.names=F,col.names=T)
28 
29 
30 print("Plotting frequency ranges plot")
31 
32 dat$best_match_class = substr(dat$best_match, 0, 3)
33 freq_labels = c("0", "0-2", "2-5", "5-10", "10-15", "15-20", "20")
34 dat$frequency_bins = cut(dat$percentage_mutations, breaks=c(-Inf, 0, 2,5,10,15,20, Inf), labels=freq_labels)
35 
36 frequency_bins_sum = data.frame(data.table(dat)[, list(class_sum=sum(.N)), by=c("best_match_class")])
37 
38 frequency_bins_data = data.frame(data.table(dat)[, list(frequency_count=.N), by=c("best_match_class", "frequency_bins")])
39 
40 frequency_bins_data = merge(frequency_bins_data, frequency_bins_sum, by="best_match_class")
41 
42 frequency_bins_data$frequency = round(frequency_bins_data$frequency_count / frequency_bins_data$class_sum * 100, 2)
43 
44 p = ggplot(frequency_bins_data, aes(frequency_bins, frequency))
45 p = p + geom_bar(aes(fill=best_match_class), stat="identity", position="dodge") + theme(panel.background = element_rect(fill = "white", colour="black"))
46 p = p + xlab("Frequency ranges") + ylab("Frequency") + ggtitle("Mutation Frequencies by class") + scale_fill_manual(values=c("IGA" = "blue4", "IGG" = "olivedrab3", "IGM" = "black"))
47 
48 png(filename="frequency_ranges.png")
49 print(p)
50 dev.off()
51 
52 frequency_bins_data_by_class = frequency_bins_data
53 
54 write.table(frequency_bins_data_by_class, "frequency_ranges_classes.txt", sep="\t",quote=F,row.names=F,col.names=T)
55 
56 frequency_bins_data = data.frame(data.table(dat)[, list(frequency_count=.N), by=c("best_match", "best_match_class", "frequency_bins")])
57 
58 frequency_bins_data = merge(frequency_bins_data, frequency_bins_sum, by="best_match_class")
59 
60 frequency_bins_data$frequency = round(frequency_bins_data$frequency_count / frequency_bins_data$class_sum * 100, 2)
61 
62 write.table(frequency_bins_data, "frequency_ranges_subclasses.txt", sep="\t",quote=F,row.names=F,col.names=T)
63 
64 
65 #frequency_bins_data_by_class
66 #frequency_ranges_subclasses.txt
67 
68 
69 
70 
71 
72 
73 
74 
75 
76 
77 
78 
79 
80 
81 
82 
83 
84 
85 
86 
87 
88 
89 
90 
91 
92 
93