annotate sequence_overview.r @ 32:4c5ba6b5d10d draft

Uploaded

1 library(reshape2)
2 
3 args <- commandArgs(trailingOnly = TRUE)
4 
5 before.unique.file = args[1]
6 merged.file = args[2]
7 outputdir = args[3]
8 gene.classes = unlist(strsplit(args[4], ","))
9 hotspot.analysis.sum.file = args[5]
10 NToverview.file = paste(outputdir, "ntoverview.txt", sep="/")
11 NTsum.file = paste(outputdir, "ntsum.txt", sep="/")
12 main.html = "index.html"
13 empty.region.filter = args[6]
14 
15 
16 setwd(outputdir)
17 
18 before.unique = read.table(before.unique.file, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
19 merged = read.table(merged.file, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
20 hotspot.analysis.sum = read.table(hotspot.analysis.sum.file, header=F, sep=",", fill=T, stringsAsFactors=F, quote="")
21 
22 #before.unique = before.unique[!grepl("unmatched", before.unique$best_match),]
23 
24 if(empty.region.filter == "leader"){
25 	before.unique$seq_conc = paste(before.unique$FR1.IMGT.seq, before.unique$CDR1.IMGT.seq, before.unique$FR2.IMGT.seq, before.unique$CDR2.IMGT.seq, before.unique$FR3.IMGT.seq, before.unique$CDR3.IMGT.seq)
26 } else if(empty.region.filter == "FR1"){
27 	before.unique$seq_conc = paste(before.unique$CDR1.IMGT.seq, before.unique$FR2.IMGT.seq, before.unique$CDR2.IMGT.seq, before.unique$FR3.IMGT.seq, before.unique$CDR3.IMGT.seq)
28 } else if(empty.region.filter == "CDR1"){
29 	before.unique$seq_conc = paste(before.unique$FR2.IMGT.seq, before.unique$CDR2.IMGT.seq, before.unique$FR3.IMGT.seq, before.unique$CDR3.IMGT.seq)
30 } else if(empty.region.filter == "FR2"){
31 	before.unique$seq_conc = paste(before.unique$CDR2.IMGT.seq, before.unique$FR3.IMGT.seq, before.unique$CDR3.IMGT.seq)
32 }
33 
34 IDs = before.unique[,c("Sequence.ID", "seq_conc", "best_match", "Functionality")]
35 IDs$best_match = as.character(IDs$best_match)
36 
37 dat = data.frame(table(before.unique$seq_conc))
38 
39 names(dat) = c("seq_conc", "Freq")
40 
41 dat$seq_conc = factor(dat$seq_conc)
42 
43 dat = dat[order(as.character(dat$seq_conc)),]
44 
45 #writing html from R...
46 get.bg.color = function(val){
47 	if(val %in% c("TRUE", "FALSE", "T", "F")){ #if its a logical value, give the background a green/red color
48 		return(ifelse(val,"#eafaf1","#f9ebea"))
49 	} else if (!is.na(as.numeric(val))) { #if its a numerical value, give it a grey tint if its >0
50 		return(ifelse(val > 0,"#eaecee","white"))
51 	} else {
52 		return("white")
53 	}
54 }
55 td = function(val) {
56   return(paste("<td bgcolor='", get.bg.color(val), "'>", val, "</td>", sep=""))
57 }
58 tr = function(val) { 
59 	return(paste(c("<tr>", sapply(val, td), "</tr>"), collapse="")) 
60 }
61 
62 make.link = function(id, clss, val) { 
63 	paste("<a href='", clss, "_", id, ".html'>", val, "</a>", sep="") 
64 }
65 tbl = function(df) {
66 	res = "<table border='1'>"
67 	for(i in 1:nrow(df)){ 
68 		res = paste(res, tr(df[i,]), sep="")
69 	}
70 	res = paste(res, "</table>")
71 }
72 
73 cat("<table border='1' class='pure-table pure-table-striped'>", file=main.html, append=F)
74 
75 if(empty.region.filter == "leader"){
76 	cat("<caption>FR1+CDR1+FR2+CDR2+FR3+CDR3 sequences that show up more than once</caption>", file=main.html, append=T)
77 } else if(empty.region.filter == "FR1"){
78 	cat("<caption>CDR1+FR2+CDR2+FR3+CDR3 sequences that show up more than once</caption>", file=main.html, append=T)
79 } else if(empty.region.filter == "CDR1"){
80 	cat("<caption>FR2+CDR2+FR3+CDR3 sequences that show up more than once</caption>", file=main.html, append=T)
81 } else if(empty.region.filter == "FR2"){
82 	cat("<caption>CDR2+FR3+CDR3 sequences that show up more than once</caption>", file=main.html, append=T)
83 }
84 
85 cat("<tr>", file=main.html, append=T)
86 cat("<th>Sequence</th><th>Functionality</th><th>IGA1</th><th>IGA2</th><th>IGG1</th><th>IGG2</th><th>IGG3</th><th>IGG4</th><th>IGM</th><th>IGE</th><th>UN</th>", file=main.html, append=T)
87 cat("<th>total IGA</th><th>total IGG</th><th>total IGE</th><th>total IGM</th><th>number of subclasses</th><th>present in both IGA and IGG</th><th>present in IGA, IGG and IGM</th><th>present in IGA, IGG and IGE</th><th>present in IGA, IGG, IGM and IGE</th><th>IGA1+IGA2</th>", file=main.html, append=T)
88 cat("<th>IGG1+IGG2</th><th>IGG1+IGG3</th><th>IGG1+IGG4</th><th>IGG2+IGG3</th><th>IGG2+IGG4</th><th>IGG3+IGG4</th>", file=main.html, append=T)
89 cat("<th>IGG1+IGG2+IGG3</th><th>IGG2+IGG3+IGG4</th><th>IGG1+IGG2+IGG4</th><th>IGG1+IGG3+IGG4</th><th>IGG1+IGG2+IGG3+IGG4</th>", file=main.html, append=T)
90 cat("</tr>", file=main.html, append=T)
91 
92 
93 
94 single.sequences=0 #sequence only found once, skipped
95 in.multiple=0 #same sequence across multiple subclasses
96 multiple.in.one=0 #same sequence multiple times in one subclass
97 unmatched=0 #all of the sequences are unmatched
98 some.unmatched=0 #one or more sequences in a clone are unmatched
99 matched=0 #should be the same als matched sequences
00 
01 sequence.id.page="by_id.html"
02 
03 for(i in 1:nrow(dat)){
04 	
05 	ca1 = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGA1", IDs$best_match),]
06 	ca2 = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGA2", IDs$best_match),]
07 	
08 	cg1 = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGG1", IDs$best_match),]
09 	cg2 = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGG2", IDs$best_match),]
10 	cg3 = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGG3", IDs$best_match),]
11 	cg4 = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGG4", IDs$best_match),]
12 	
13 	cm = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGM", IDs$best_match),]
14 	
15 	ce = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^IGE", IDs$best_match),]
16 	
17 	un = IDs[IDs$seq_conc == dat[i,c("seq_conc")] & grepl("^unmatched", IDs$best_match),]
18 	
19 	allc = rbind(ca1, ca2, cg1, cg2, cg3, cg4, cm, ce, un)
20 	
21 	ca1.n = nrow(ca1)
22 	ca2.n = nrow(ca2)
23 	
24 	cg1.n = nrow(cg1)
25 	cg2.n = nrow(cg2)
26 	cg3.n = nrow(cg3)
27 	cg4.n = nrow(cg4)
28 	
29 	cm.n = nrow(cm)
30 	
31 	ce.n = nrow(ce)
32 	
33 	un.n = nrow(un)
34 	
35 	classes = c(ca1.n, ca2.n, cg1.n, cg2.n, cg3.n, cg4.n, cm.n, ce.n, un.n)
36 	
37 	classes.sum = sum(classes)
38 	
39 	if(classes.sum == 1){
40 		single.sequences = single.sequences + 1
41 		next
42 	}
43 	
44 	if(un.n == classes.sum){
45 		unmatched = unmatched + 1
46 		next
47 	}
48 	
49 	in.classes = sum(classes > 0)
50 	
51 	matched = matched + in.classes #count in how many subclasses the sequence occurs.
52 	
53 	if(any(classes == classes.sum)){
54 		multiple.in.one = multiple.in.one + 1
55 	} else if (un.n > 0) {
56 		some.unmatched = some.unmatched + 1
57 	} else {
58 		in.multiple = in.multiple + 1
59 	}
60 	
61 	id = as.numeric(dat[i,"seq_conc"])
62 	
63 	functionality = paste(unique(allc[,"Functionality"]), collapse=",")
64 	
65 	by.id.row = c()
66 	
67 	if(ca1.n > 0){
68 		cat(tbl(ca1), file=paste("IGA1_", id, ".html", sep=""))
69 	}
70 
71 	if(ca2.n > 0){
72 		cat(tbl(ca2), file=paste("IGA2_", id, ".html", sep=""))
73 	}
74 
75 	if(cg1.n > 0){
76 		cat(tbl(cg1), file=paste("IGG1_", id, ".html", sep=""))
77 	}
78 
79 	if(cg2.n > 0){
80 		cat(tbl(cg2), file=paste("IGG2_", id, ".html", sep=""))
81 	}
82 
83 	if(cg3.n > 0){
84 		cat(tbl(cg3), file=paste("IGG3_", id, ".html", sep=""))
85 	}
86 
87 	if(cg4.n > 0){
88 		cat(tbl(cg4), file=paste("IGG4_", id, ".html", sep=""))
89 	}
90 
91 	if(cm.n > 0){
92 		cat(tbl(cm), file=paste("IGM_", id, ".html", sep=""))
93 	}
94 
95 	if(ce.n > 0){
96 		cat(tbl(ce), file=paste("IGE_", id, ".html", sep=""))
97 	}
98 
99 	if(un.n > 0){
00 		cat(tbl(un), file=paste("un_", id, ".html", sep=""))
01 	}
02 	
03 	ca1.html = make.link(id, "IGA1", ca1.n)
04 	ca2.html = make.link(id, "IGA2", ca2.n)
05 	
06 	cg1.html = make.link(id, "IGG1", cg1.n)
07 	cg2.html = make.link(id, "IGG2", cg2.n)
08 	cg3.html = make.link(id, "IGG3", cg3.n)
09 	cg4.html = make.link(id, "IGG4", cg4.n)
10 	
11 	cm.html = make.link(id, "IGM", cm.n)
12 	
13 	ce.html = make.link(id, "IGE", ce.n)
14 	
15 	un.html = make.link(id, "un", un.n)
16 	
17 	#extra columns
18 	ca.n = ca1.n + ca2.n
19 	
20 	cg.n = cg1.n + cg2.n + cg3.n + cg4.n
21 	
22 	#in.classes
23 	
24 	in.ca.cg = (ca.n > 0 & cg.n > 0)
25 	
26 	in.ca.cg.cm = (ca.n > 0 & cg.n > 0 & cm.n > 0)
27 	
28 	in.ca.cg.ce = (ca.n > 0 & cg.n > 0 & ce.n > 0)
29 	
30 	in.ca.cg.cm.ce = (ca.n > 0 & cg.n > 0 & cm.n > 0 & ce.n > 0)
31 	
32 	in.ca1.ca2 = (ca1.n > 0 & ca2.n > 0)
33 	
34 	in.cg1.cg2 = (cg1.n > 0 & cg2.n > 0)
35 	in.cg1.cg3 = (cg1.n > 0 & cg3.n > 0)
36 	in.cg1.cg4 = (cg1.n > 0 & cg4.n > 0)
37 	in.cg2.cg3 = (cg2.n > 0 & cg3.n > 0)
38 	in.cg2.cg4 = (cg2.n > 0 & cg4.n > 0)
39 	in.cg3.cg4 = (cg3.n > 0 & cg4.n > 0)
40 	
41 	in.cg1.cg2.cg3 = (cg1.n > 0 & cg2.n > 0 & cg3.n > 0)
42 	in.cg2.cg3.cg4 = (cg2.n > 0 & cg3.n > 0 & cg4.n > 0)
43 	in.cg1.cg2.cg4 = (cg1.n > 0 & cg2.n > 0 & cg4.n > 0)
44 	in.cg1.cg3.cg4 = (cg1.n > 0 & cg3.n > 0 & cg4.n > 0)
45 	
46 	in.cg.all = (cg1.n > 0 & cg2.n > 0 & cg3.n > 0 & cg4.n > 0)
47 	
48 	
49 	
50 	
51 	#rw = c(as.character(dat[i,"seq_conc"]), functionality, ca1.html, ca2.html, cg1.html, cg2.html, cg3.html, cg4.html, cm.html, un.html)
52 	rw = c(as.character(dat[i,"seq_conc"]), functionality, ca1.html, ca2.html, cg1.html, cg2.html, cg3.html, cg4.html, cm.html, ce.html, un.html)
53 	rw = c(rw, ca.n, cg.n, cm.n, ce.n, in.classes, in.ca.cg, in.ca.cg.cm, in.ca.cg.ce, in.ca.cg.cm.ce, in.ca1.ca2, in.cg1.cg2, in.cg1.cg3, in.cg1.cg4, in.cg2.cg3, in.cg2.cg4, in.cg3.cg4, in.cg1.cg2.cg3, in.cg2.cg3.cg4, in.cg1.cg2.cg4, in.cg1.cg3.cg4, in.cg.all)
54 
55 	cat(tr(rw), file=main.html, append=T)
56 	
57 	
58 	for(i in 1:nrow(allc)){ #generate html by id
59 		html = make.link(id, allc[i,"best_match"], allc[i,"Sequence.ID"])
60 		cat(paste(html, "<br />"), file=sequence.id.page, append=T)
61 	}
62 }
63 
64 cat("</table>", file=main.html, append=T)
65 
66 print(paste("Single sequences:", single.sequences))
67 print(paste("Sequences in multiple subclasses:", in.multiple))
68 print(paste("Multiple sequences in one subclass:", multiple.in.one))
69 print(paste("Matched with unmatched:", some.unmatched))
70 print(paste("Count that should match 'matched' sequences:", matched))
71 
72 #ACGT overview
73 
74 #NToverview = merged[!grepl("^unmatched", merged$best_match),]
75 NToverview = merged
76 
77 if(empty.region.filter == "leader"){
78 	NToverview$seq = paste(NToverview$FR1.IMGT.seq, NToverview$CDR1.IMGT.seq, NToverview$FR2.IMGT.seq, NToverview$CDR2.IMGT.seq, NToverview$FR3.IMGT.seq)
79 } else if(empty.region.filter == "FR1"){
80 	NToverview$seq = paste(NToverview$CDR1.IMGT.seq, NToverview$FR2.IMGT.seq, NToverview$CDR2.IMGT.seq, NToverview$FR3.IMGT.seq)
81 } else if(empty.region.filter == "CDR1"){
82 	NToverview$seq = paste(NToverview$FR2.IMGT.seq, NToverview$CDR2.IMGT.seq, NToverview$FR3.IMGT.seq)
83 } else if(empty.region.filter == "FR2"){
84 	NToverview$seq = paste(NToverview$CDR2.IMGT.seq, NToverview$FR3.IMGT.seq)
85 }
86 
87 NToverview$A = nchar(gsub("[^Aa]", "", NToverview$seq))
88 NToverview$C = nchar(gsub("[^Cc]", "", NToverview$seq))
89 NToverview$G = nchar(gsub("[^Gg]", "", NToverview$seq))
90 NToverview$T = nchar(gsub("[^Tt]", "", NToverview$seq))
91 
92 #Nsum = data.frame(Sequence.ID="-", best_match="Sum", seq="-", A = sum(NToverview$A), C = sum(NToverview$C), G = sum(NToverview$G), T = sum(NToverview$T))
93 
94 #NToverview = rbind(NToverview, NTsum)
95 
96 NTresult = data.frame(nt=c("A", "C", "T", "G"))
97 
98 for(clazz in gene.classes){
99 	print(paste("class:", clazz))
00 	NToverview.sub = NToverview[grepl(paste("^", clazz, sep=""), NToverview$best_match),]
01 	print(paste("nrow:", nrow(NToverview.sub)))
02 	new.col.x = c(sum(NToverview.sub$A), sum(NToverview.sub$C), sum(NToverview.sub$T), sum(NToverview.sub$G))
03 	new.col.y = sum(new.col.x)
04 	new.col.z = round(new.col.x / new.col.y * 100, 2)
05 	
06 	tmp = names(NTresult)
07 	NTresult = cbind(NTresult, data.frame(new.col.x, new.col.y, new.col.z))
08 	names(NTresult) = c(tmp, paste(clazz, c("x", "y", "z"), sep=""))
09 }
10 
11 write.table(NToverview[,c("Sequence.ID", "best_match", "seq", "A", "C", "G", "T")], NToverview.file, quote=F, sep="\t", row.names=F, col.names=T)
12 
13 NToverview = NToverview[!grepl("unmatched", NToverview$best_match),]
14 
15 new.col.x = c(sum(NToverview$A), sum(NToverview$C), sum(NToverview$T), sum(NToverview$G))
16 new.col.y = sum(new.col.x)
17 new.col.z = round(new.col.x / new.col.y * 100, 2)
18 
19 tmp = names(NTresult)
20 NTresult = cbind(NTresult, data.frame(new.col.x, new.col.y, new.col.z))
21 names(NTresult) = c(tmp, paste("all", c("x", "y", "z"), sep=""))
22 
23 names(hotspot.analysis.sum) = names(NTresult)
24 
25 hotspot.analysis.sum = rbind(hotspot.analysis.sum, NTresult)
26 
27 write.table(hotspot.analysis.sum, hotspot.analysis.sum.file, quote=F, sep=",", row.names=F, col.names=F, na="0")
28 
29 
30 
31 
32 
33 
34 
35 
36 
37 
38 
39 
40 
41 
42 
43 
44 
45 
46 
47 
48 
49 
50 
51 
52 
53 
54 
55 
56 
57