0
|
1 #!/usr/bin/perl -w
|
|
2 use strict;
|
|
3
|
|
4 #convert from master variant file to snp table (Webb format?)
|
|
5 #new format for version 2.0, also different format for cancer normal pairs
|
|
6 #set columns for 2.0 version Cancer format
|
|
7 my $aCnt1 = 21;
|
|
8 my $aCnt2 = 22;
|
|
9
|
|
10 #snp table format:
|
|
11 #1. chr
|
|
12 #2. position (0 based)
|
|
13 #3. ref allele
|
|
14 #4. second allele
|
|
15 #5. overall quality
|
|
16 #foreach individual (6-9, 10-13, ...)
|
|
17 #a. count of allele in 3
|
|
18 #b. count of allele in 4
|
|
19 #c. genotype call (-1, or count of ref allele)
|
|
20 #d. quality of genotype call (quality of non-ref allele from masterVar)
|
|
21
|
|
22 if (!@ARGV) {
|
|
23 print "usage: master2gd_snp.pl masterVar.txt[.gz|.bz2] [-tab=snpTable.txt -addColsOnly -build=hg19 -name=na ] > newSnpTable.txt\n";
|
|
24 exit;
|
|
25 }
|
|
26
|
|
27 my $in = shift @ARGV;
|
|
28 my $tab;
|
|
29 my $tabOnly;
|
|
30 my $build;
|
|
31 my $name;
|
|
32 foreach (@ARGV) {
|
|
33 if (/-tab=(.*)/) { $tab = $1; }
|
|
34 elsif (/-addColsOnly/) { $tabOnly = 1; }
|
|
35 elsif (/-build=(.*)/) { $build = $1; }
|
|
36 elsif (/-name=(.*)/) { $name = $1; }
|
|
37 }
|
|
38
|
|
39 #WARNING loads snp table in memory, this could take > 1G ram
|
|
40 my %old;
|
|
41 my $colcnt = 0;
|
|
42 my @head;
|
|
43 if ($tab) {
|
|
44 open(FH, $tab) or die "Couldn't open $tab, $!\n";
|
|
45 while (<FH>) {
|
|
46 chomp;
|
|
47 if (/^#/) { push(@head, $_); next; }
|
|
48 my @f = split(/\t/);
|
|
49 $old{"$f[0]:$f[1]"} = join("\t", @f);
|
|
50 $colcnt = scalar @f;
|
|
51 }
|
|
52 close FH or die "Couldn't close $tab, $!\n";
|
|
53 }
|
|
54
|
|
55 if ($in =~ /.gz$/) {
|
|
56 open(FH, "zcat $in |") or die "Couldn't open $in, $!\n";
|
|
57 }elsif ($in =~ /.bz2$/) {
|
|
58 open(FH, "bzcat $in |") or die "Couldn't open $in, $!\n";
|
|
59 }else {
|
|
60 open(FH, $in) or die "Couldn't open $in, $!\n";
|
|
61 }
|
|
62 prepHeader();
|
|
63 if (@head) { #keep old header, add new?
|
|
64 print join("\n", @head), "\n";
|
|
65 }
|
|
66 while (<FH>) {
|
|
67 chomp;
|
|
68 #FORMAT_VERSION 2.0
|
|
69 if (/^#FORMAT_VERSION\s+1\./) {
|
|
70 $aCnt1 = 16;
|
|
71 $aCnt2 = 17;
|
|
72 }
|
|
73 if (/^#/) { next; }
|
|
74 if (/^>/) { next; } #headers
|
|
75 if (/^\s*$/) { next; }
|
|
76 my @f = split(/\t/);
|
|
77 if (!$f[6]) { next; } #WHAT? most likely still zipped?
|
|
78 if ($f[6] ne 'snp') { next; } #table only has substitutions
|
|
79 if ($f[5] eq 'het-alt') { next; } #skip heterozygous with no ref match
|
|
80 if ($f[5] =~ /(hom|het)/) { #zygosity #haploid chrX and chrY?
|
|
81 my $a = $f[7]; #reference allele
|
|
82 my $a2;
|
|
83 my $freq;
|
|
84 my $freq2;
|
|
85 my $sc = -1;
|
|
86 my $alt;
|
|
87 my $g = 1; #genotype == ref allele count
|
|
88 if ($f[8] eq $f[9]) { #should be homozygous?
|
|
89 $a2 = $f[8];
|
|
90 $g = 0;
|
|
91 if ($f[10] && $f[10] ne '') { $sc = $f[10]; }#is this the best one to use? or smallest?
|
|
92 }else {
|
|
93 if ($a ne $f[8]) { $a2 = $f[8]; $alt = 8; }
|
|
94 elsif ($a ne $f[9]) { $a2 = $f[9]; $alt = 9; }
|
|
95 }
|
|
96 if (defined $f[10] && defined $f[11] && $alt) { #VAF score in 2.0 format
|
|
97 if ($f[$alt+2] && $f[$alt+2] ne '') { $sc = $f[$alt+2]; }
|
|
98 }
|
|
99 #version 1.12 columns 16 & 17, version 2.0 Cancer columns 21 & 22
|
|
100 if (defined $f[$aCnt1] && defined $f[$aCnt2] && $alt) {
|
|
101 if ($alt == 8) {
|
|
102 $freq = $f[$aCnt2];
|
|
103 $freq2 = $f[$aCnt1];
|
|
104 }elsif ($alt == 9) {
|
|
105 $freq = $f[$aCnt1];
|
|
106 $freq2 = $f[$aCnt2];
|
|
107 }
|
|
108 }elsif (defined $f[$aCnt1]) {
|
|
109 $freq = 0;
|
|
110 $freq2 = $f[$aCnt1];
|
|
111 }
|
|
112 #if starting a new table or new SNP in old table
|
|
113 #add option to only build on current table?
|
|
114 if (!$tab) {
|
|
115 print "$f[2]\t$f[3]\t$a\t$a2\t-1";
|
|
116 }elsif (!$tabOnly && !exists $old{"$f[2]:$f[3]"}) {
|
|
117 print "$f[2]\t$f[3]\t$a\t$a2\t-1";
|
|
118 }elsif (exists $old{"$f[2]:$f[3]"}) {
|
|
119 print $old{"$f[2]:$f[3]"};
|
|
120 $old{"$f[2]:$f[3]"} = ''; #unset so we know it is printed
|
|
121 }elsif ($tabOnly && !exists $old{"$f[2]:$f[3]"}) {
|
|
122 next; #skip this one entirely
|
|
123 }
|
|
124 if ($colcnt && !exists $old{"$f[2]:$f[3]"}) {
|
|
125 #new SNP pad for missing individuals
|
|
126 my $i = 5;
|
|
127 while ($i < $colcnt) {
|
|
128 print "\t-1\t-1\t-1\t-1";
|
|
129 $i += 4;
|
|
130 }
|
|
131 }
|
|
132 #add columns for individual
|
|
133 print "\t$freq\t$freq2\t$g\t$sc\n";
|
|
134 }elsif ($f[5] eq 'hap') {
|
|
135 my $g = 0;
|
|
136 my $freq = 0;
|
|
137 my $freq2 = 0;
|
|
138 if (defined $f[10]) { $freq2 = $f[10]; }
|
|
139 my $sc = -1;
|
|
140 if (defined $f[$aCnt1] && $f[$aCnt1] ne '') { $sc = $f[$aCnt1]; }
|
|
141 if ($f[8]) {
|
|
142 if (!$tab) {
|
|
143 print "$f[2]\t$f[3]\t$f[7]\t$f[8]\t-1";
|
|
144 }elsif (!$tabOnly && !exists $old{"$f[2]:$f[3]"}) {
|
|
145 print "$f[2]\t$f[3]\t$f[7]\t$f[8]\t-1";
|
|
146 }elsif (exists $old{"$f[2]:$f[3]"}) {
|
|
147 print $old{"$f[2]:$f[3]"};
|
|
148 $old{"$f[2]:$f[3]"} = ''; #unset so we know it is printed
|
|
149 }elsif ($tabOnly && !exists $old{"$f[2]:$f[3]"}) {
|
|
150 next; #skip this one entirely
|
|
151 }
|
|
152 if ($colcnt && !exists $old{"$f[2]:$f[3]"}) {
|
|
153 #new SNP pad for missing individuals
|
|
154 my $i = 5;
|
|
155 while ($i < $colcnt) {
|
|
156 print "\t-1\t-1\t-1\t-1";
|
|
157 $i += 4;
|
|
158 }
|
|
159 }
|
|
160 #add columns for individual
|
|
161 print "\t$freq\t$freq2\t$g\t$sc\n";
|
|
162 }
|
|
163 }
|
|
164 }
|
|
165 close FH or die "Couldn't close $in, $!\n";
|
|
166
|
|
167 #if adding to a snp table, now we need to finish those not in the latest set
|
|
168 foreach my $k (keys %old) {
|
|
169 if ($old{$k} ne '') { #not printed yet
|
|
170 print $old{$k}, "\t-1\t-1\t-1\t-1\n"; #plus blank for this one
|
|
171 }
|
|
172 }
|
|
173
|
|
174 exit;
|
|
175
|
|
176 #parse old header and add or create new
|
|
177 sub prepHeader {
|
|
178 if (!$build) { $build = 'hg19'; } #set default
|
|
179 my @cnames;
|
|
180 my @ind;
|
|
181 my $n;
|
|
182 if (@head) { #parse previous header
|
|
183 my $h = join("", @head); #may split between lines
|
|
184 if ($h =~ /"column_names":\[(.*?)\]/) {
|
|
185 my @t = split(/,/, $1);
|
|
186 foreach (@t) { s/"//g; }
|
|
187 @cnames = @t;
|
|
188 $n = $cnames[$#cnames];
|
|
189 $n =~ s/Q//;
|
|
190 $n++;
|
|
191 }
|
|
192 if ($h =~ /"dbkey":"(.*?)"/) { $build = $1; }
|
|
193 if ($h =~ /"individuals":\[(.*)\]/) {
|
|
194 my $t = $1;
|
|
195 $t =~ s/\]\].*/]/; #remove if there is more categories
|
|
196 @ind = split(/,/, $t);
|
|
197 }
|
|
198 }else { #start new header
|
|
199 @cnames = ("chr", "pos", "A", "B", "Q");
|
|
200 $n = 1;
|
|
201 }
|
|
202 #add current
|
|
203 if (!$name) { $name= 'na'; }
|
|
204 my $stcol = $colcnt + 1;
|
|
205 if ($stcol == 1) { $stcol = 6; } #move past initial columns
|
|
206 push(@ind, "[\"$name\",$stcol]");
|
|
207 push(@cnames, "${n}A", "${n}B", "${n}G", "${n}Q");
|
|
208 #reassign head
|
|
209 undef @head;
|
|
210 foreach (@cnames) { $_ = "\"$_\""; } #quote name
|
|
211 $head[0] = "#{\"column_names\":[" . join(",", @cnames) . "],";
|
|
212 $head[1] = "#\"individuals\":[" . join(",", @ind) . "],";
|
|
213 $head[2] = "#\"dbkey\":\"$build\",\"pos\":2,\"rPos\":2,\"ref\":1,\"scaffold\":1,\"species\":\"$build\"}";
|
|
214 }
|
|
215 ####End
|
|
216
|
|
217 ##example header
|
|
218 #{"column_names":["chr","pos","A","B","Q","1A","1B","1G","1Q","2A","2B","2G","2Q","3A","3B","3G","3Q","4A","4B","4G","4Q","5A","5B","5G","5Q","6A","6B","6G","6Q","7A","7B","7G","7Q","8A","8B","8G",
|
|
219 #"8Q","9A","9B","9G","9Q","10A","10B","10G","10Q"],"dbkey":"hg19","individuals":[["Boh_15M",6],["Boh_19M",10],["Paya_27F",14],["Paya_2F",18],["Paya_32F",22],["Ruil_2M",26],["Ruil_36M",30],["Ruil_3M",
|
|
220 #34],["Ruil_40",38],["Ruil_47F",42]],"pos":2,"rPos":2,"ref":1,"scaffold":1,"species":"hg19"}
|
|
221 #chr1 10290 C T 46.4 0 2 0 7 1 2 0 4 3 2 1 22 0 0 -1 0 1 0 1 4 0 2 0 7 0 0 -1 0 2 3 1 14 0 1 0 4 1 1 1 6
|