Mercurial > repos > davidvanzessen > shm_csr

comparison merge_and_filter.r @ 0:c33d93683a09 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author davidvanzessen
date Thu, 13 Oct 2016 10:52:24 -0400
parents
children faae21ba5c63
comparison
equal deleted inserted replaced
-1:000000000000 0:c33d93683a09
1 args <- commandArgs(trailingOnly = TRUE)
2
3
4 summaryfile = args[1]
5 sequencesfile = args[2]
6 mutationanalysisfile = args[3]
7 mutationstatsfile = args[4]
8 hotspotsfile = args[5]
9 gene_identification_file= args[6]
10 output = args[7]
11 before.unique.file = args[8]
12 unmatchedfile = args[9]
13 method=args[10]
14 functionality=args[11]
15 unique.type=args[12]
16 filter.unique=args[13]
17 class.filter=args[14]
18 empty.region.filter=args[15]
19
20 summ = read.table(summaryfile, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
21 sequences = read.table(sequencesfile, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
22 mutationanalysis = read.table(mutationanalysisfile, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
23 mutationstats = read.table(mutationstatsfile, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
24 hotspots = read.table(hotspotsfile, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
25 gene_identification = read.table(gene_identification_file, header=T, sep="\t", fill=T, stringsAsFactors=F, quote="")
26
27 if(method == "blastn"){
28 "qseqid\tsseqid\tpident\tlength\tmismatch\tgapopen\tqstart\tqend\tsstart\tsend\tevalue\tbitscore"
29 gene_identification = gene_identification[!duplicated(gene_identification$qseqid),]
30 ref_length = data.frame(sseqid=c("ca1", "ca2", "cg1", "cg2", "cg3", "cg4", "cm"), ref.length=c(81,81,141,141,141,141,52))
31 gene_identification = merge(gene_identification, ref_length, by="sseqid", all.x=T)
32 gene_identification$chunk_hit_percentage = (gene_identification$length / gene_identification$ref.length) * 100
33 gene_identification = gene_identification[,c("qseqid", "chunk_hit_percentage", "pident", "qstart", "sseqid")]
34 colnames(gene_identification) = c("Sequence.ID", "chunk_hit_percentage", "nt_hit_percentage", "start_locations", "best_match")
35
36 }
37
38 input.sequence.count = nrow(summ)
39 print(paste("Number of sequences in summary file:", input.sequence.count))
40
41 filtering.steps = data.frame(character(0), numeric(0))
42
43 filtering.steps = rbind(filtering.steps, c("Input", input.sequence.count))
44
45 filtering.steps[,1] = as.character(filtering.steps[,1])
46 filtering.steps[,2] = as.character(filtering.steps[,2])
47 #filtering.steps[,3] = as.numeric(filtering.steps[,3])
48
49 summ = merge(summ, gene_identification, by="Sequence.ID")
50
51 summ = summ[summ$Functionality != "No results",]
52
53 print(paste("Number of sequences after 'No results' filter:", nrow(summ)))
54
55 filtering.steps = rbind(filtering.steps, c("After 'No results' filter", nrow(summ)))
56
57 if(functionality == "productive"){
58 summ = summ[summ$Functionality == "productive (see comment)" | summ$Functionality == "productive",]
59 } else if (functionality == "unproductive"){
60 summ = summ[summ$Functionality == "unproductive (see comment)" | summ$Functionality == "unproductive",]
61 } else if (functionality == "remove_unknown"){
62 summ = summ[summ$Functionality != "No results" & summ$Functionality != "unknown (see comment)" & summ$Functionality != "unknown",]
63 }
64
65 print(paste("Number of sequences after productive filter:", nrow(summ)))
66
67 filtering.steps = rbind(filtering.steps, c("After productive filter", nrow(summ)))
68
69 splt = strsplit(class.filter, "_")[[1]]
70 chunk_hit_threshold = as.numeric(splt[1])
71 nt_hit_threshold = as.numeric(splt[2])
72
73 higher_than=(summ$chunk_hit_percentage >= chunk_hit_threshold & summ$nt_hit_percentage >= nt_hit_threshold)
74
75 unmatched=summ[NULL,c("Sequence.ID", "chunk_hit_percentage", "nt_hit_percentage", "start_locations", "best_match")]
76
77 if(!all(higher_than, na.rm=T)){ #check for 'not all' because that would mean the unmatched set is empty
78 unmatched = summ[!higher_than,]
79 unmatched = unmatched[,c("Sequence.ID", "chunk_hit_percentage", "nt_hit_percentage", "start_locations", "best_match")]
80 unmatched$best_match = paste("unmatched,", unmatched$best_match)
81 summ[!higher_than,"best_match"] = paste("unmatched,", summ[!higher_than,"best_match"])
82 }
83
84 if(any(higher_than, na.rm=T)){
85 #summ = summ[higher_than,]
86 }
87
88 if(nrow(summ) == 0){
89 stop("No data remaining after filter")
90 }
91
92 result = merge(summ, mutationanalysis[,!(names(mutationanalysis) %in% names(summ)[-1])], by="Sequence.ID")
93
94 print(paste("Number of sequences after merging with mutation analysis file:", nrow(result)))
95
96 result = merge(result, mutationstats[,!(names(mutationstats) %in% names(result)[-1])], by="Sequence.ID")
97
98 print(paste("Number of sequences after merging with mutation stats file:", nrow(result)))
99
100 result = merge(result, hotspots[,!(names(hotspots) %in% names(result)[-1])], by="Sequence.ID")
101
102 print(paste("Number of sequences after merging with hotspots file:", nrow(result)))
103
104 sequences = sequences[,c("Sequence.ID", "FR1.IMGT", "CDR1.IMGT", "FR2.IMGT", "CDR2.IMGT", "FR3.IMGT", "CDR3.IMGT")]
105 names(sequences) = c("Sequence.ID", "FR1.IMGT.seq", "CDR1.IMGT.seq", "FR2.IMGT.seq", "CDR2.IMGT.seq", "FR3.IMGT.seq", "CDR3.IMGT.seq")
106 result = merge(result, sequences, by="Sequence.ID", all.x=T)
107
108 print(paste("Number of sequences in result after merging with sequences:", nrow(result)))
109
110 result$VGene = gsub("^Homsap ", "", result$V.GENE.and.allele)
111 result$VGene = gsub("[*].*", "", result$VGene)
112 result$DGene = gsub("^Homsap ", "", result$D.GENE.and.allele)
113 result$DGene = gsub("[*].*", "", result$DGene)
114 result$JGene = gsub("^Homsap ", "", result$J.GENE.and.allele)
115 result$JGene = gsub("[*].*", "", result$JGene)
116
117 result$past = do.call(paste, c(result[unlist(strsplit(unique.type, ","))], sep = ":"))
118
119 result = result[!(duplicated(result$past)), ]
120
121 result = result[,!(names(result) %in% c("past"))]
122
123 print(paste("Number of sequences in result after", unique.type, "filtering:", nrow(result)))
124
125 filtering.steps = rbind(filtering.steps, c("After duplicate filter", nrow(result)))
126
127 print(paste("Number of empty CDR1 sequences:", sum(result$CDR1.IMGT.seq == "")))
128 print(paste("Number of empty FR2 sequences:", sum(result$FR2.IMGT.seq == "")))
129 print(paste("Number of empty CDR2 sequences:", sum(result$CDR2.IMGT.seq == "")))
130 print(paste("Number of empty FR3 sequences:", sum(result$FR3.IMGT.seq == "")))
131
132 if(empty.region.filter == "FR1"){
133 result = result[result$CDR1.IMGT.seq != "" & result$FR2.IMGT.seq != "" & result$CDR2.IMGT.seq != "" & result$FR3.IMGT.seq != "", ]
134 print(paste("Number of sequences after empty CDR1, FR2, CDR2 and FR3 column filter:", nrow(result)))
135 filtering.steps = rbind(filtering.steps, c("After empty CDR1, FR2, CDR2, FR3 filter", nrow(result)))
136 } else if(empty.region.filter == "CDR1"){
137 result = result[result$FR2.IMGT.seq != "" & result$CDR2.IMGT.seq != "" & result$FR3.IMGT.seq != "", ]
138 print(paste("Number of sequences after empty FR2, CDR2 and FR3 column filter:", nrow(result)))
139 filtering.steps = rbind(filtering.steps, c("After empty FR2, CDR2, FR3 filter", nrow(result)))
140 } else if(empty.region.filter == "FR2"){
141 result = result[result$CDR2.IMGT.seq != "" & result$FR3.IMGT.seq != "", ]
142 print(paste("Number of sequences after empty CDR2 and FR3 column filter:", nrow(result)))
143 filtering.steps = rbind(filtering.steps, c("After empty CDR2, FR3 filter", nrow(result)))
144 }
145
146 if(empty.region.filter == "FR1"){
147 result = result[!(grepl("n|N", result$FR2.IMGT.seq) | grepl("n|N", result$FR3.IMGT.seq) | grepl("n|N", result$CDR1.IMGT.seq) | grepl("n|N", result$CDR2.IMGT.seq) | grepl("n|N", result$CDR3.IMGT.seq)),]
148 } else if(empty.region.filter == "CDR1"){
149 result = result[!(grepl("n|N", result$FR2.IMGT.seq) | grepl("n|N", result$FR3.IMGT.seq) | grepl("n|N", result$CDR2.IMGT.seq) | grepl("n|N", result$CDR3.IMGT.seq)),]
150 } else if(empty.region.filter == "FR2"){
151 result = result[!(grepl("n|N", result$FR3.IMGT.seq) | grepl("n|N", result$CDR3.IMGT.seq)),]
152 }
153
154 print(paste("Number of sequences in result after n filtering:", nrow(result)))
155 filtering.steps = rbind(filtering.steps, c("After N filter", nrow(result)))
156
157 cleanup_columns = c("FR1.IMGT.Nb.of.mutations",
158 "CDR1.IMGT.Nb.of.mutations",
159 "FR2.IMGT.Nb.of.mutations",
160 "CDR2.IMGT.Nb.of.mutations",
161 "FR3.IMGT.Nb.of.mutations")
162
163 for(col in cleanup_columns){
164 result[,col] = gsub("\\(.*\\)", "", result[,col])
165 result[,col] = as.numeric(result[,col])
166 result[is.na(result[,col]),] = 0
167 }
168
169 write.table(result, before.unique.file, sep="\t", quote=F,row.names=F,col.names=T)
170
171 if(filter.unique != "no"){
172 clmns = names(result)
173
174 if(empty.region.filter == "FR1"){
175 result$unique.def = paste(result$CDR1.IMGT.seq, result$FR2.IMGT.seq, result$CDR2.IMGT.seq, result$FR3.IMGT.seq, result$CDR3.IMGT.seq)
176 } else if(empty.region.filter == "CDR1"){
177 rresult$unique.def = paste(result$FR2.IMGT.seq, result$CDR2.IMGT.seq, result$FR3.IMGT.seq, result$CDR3.IMGT.seq)
178 } else if(empty.region.filter == "FR2"){
179 result$unique.def = paste(result$CDR2.IMGT.seq, result$FR3.IMGT.seq, result$CDR3.IMGT.seq)
180 }
181
182 if(grepl("_c", filter.unique)){
183 result$unique.def = paste(result$unique.def, result$best_match)
184 }
185
186 #fltr = result$unique.def %in% result.filtered$unique.def
187
188 if(grepl("keep", filter.unique)){
189 result$unique.def = paste(result$unique.def, result$best_match) #keep the unique sequences that are in multiple classes
190 result = result[!duplicated(result$unique.def),]
191 } else {
192 result = result[duplicated(result$unique.def) | duplicated(result$unique.def, fromLast=T),]
193 result$unique.def = paste(result$unique.def, result$best_match) #keep the unique sequences that are in multiple classes
194 result = result[!duplicated(result$unique.def),]
195 }
196
197 #result = result[,clmns]
198
199 #write.table(inputdata.removed, "unique_removed.csv", sep=",",quote=F,row.names=F,col.names=T)
200 }
201
202 print(paste("Number of sequences in result after CDR/FR filtering:", nrow(result)))
203 print(paste("Number of matched sequences in result after CDR/FR filtering:", nrow(result[!grepl("unmatched", result$best_match),])))
204
205 filtering.steps = rbind(filtering.steps, c("After unique filter", nrow(result)))
206
207 print(paste("Number of rows in result:", nrow(result)))
208 print(paste("Number of rows in unmatched:", nrow(unmatched)))
209
210 matched.sequences = result[!grepl("^unmatched", result$best_match),]
211
212 write.table(x=matched.sequences, file=gsub("merged.txt$", "filtered.txt", output), sep="\t",quote=F,row.names=F,col.names=T)
213
214 matched.sequences.count = nrow(matched.sequences)
215 unmatched.sequences.count = sum(grepl("^unmatched", result$best_match))
216
217 filtering.steps = rbind(filtering.steps, c("Number of matched sequences", matched.sequences.count))
218 filtering.steps = rbind(filtering.steps, c("Number of unmatched sequences", unmatched.sequences.count))
219 filtering.steps[,2] = as.numeric(filtering.steps[,2])
220 filtering.steps$perc = round(filtering.steps[,2] / input.sequence.count * 100, 2)
221
222 write.table(x=filtering.steps, file=gsub("unmatched", "filtering_steps", unmatchedfile), sep="\t",quote=F,row.names=F,col.names=F)
223
224 write.table(x=result, file=output, sep="\t",quote=F,row.names=F,col.names=T)
225 write.table(x=unmatched, file=unmatchedfile, sep="\t",quote=F,row.names=F,col.names=T)