3 # VCF specs: http://www.1000genomes.org/wiki/doku.php?id=1000_genomes:analysis:vcf3.3
12 my $opts = parse_params();
13 do_pileup_to_vcf($opts);
22 if ( scalar @msg ) { croak(@msg); }
24 "Usage: sam2vcf.pl [OPTIONS] < in.pileup > out.vcf\n",
26 " -i, --indels-only Ignore SNPs.\n",
27 " -r, --refseq <file.fa> The reference sequence, required when indels are present.\n",
28 " -s, --snps-only Ignore indels.\n",
29 " -t, --column-title <string> The column title.\n",
30 " -h, -?, --help This help message.\n",
39 $opts{fh_in} = *STDIN;
40 $opts{fh_out} = *STDOUT;
42 while (my $arg=shift(@ARGV))
44 if ( $arg eq '-r' || $arg eq '--refseq' ) { $opts{refseq}=shift(@ARGV); next; }
45 if ( $arg eq '-t' || $arg eq '--column-title' ) { $opts{title}=shift(@ARGV); next; }
46 if ( $arg eq '-s' || $arg eq '--snps-only' ) { $opts{snps_only}=1; next; }
47 if ( $arg eq '-i' || $arg eq '--indels-only' ) { $opts{indels_only}=1; next; }
48 if ( $arg eq '-?' || $arg eq '-h' || $arg eq '--help' ) { error(); }
50 error("Unknown parameter \"$arg\". Run -h for help.\n");
66 if ( !exists($iupac{$base}) )
68 if ( $base ne 'A' && $base ne 'C' && $base ne 'G' && $base ne 'T' ) { error("FIXME: what is this [$base]?\n"); }
69 if ( $ref eq $base ) { return ('.','0/0'); }
72 my $gt = $iupac{$base};
73 if ( $$gt[0] eq $ref ) { return ($$gt[1],'0/1'); }
74 elsif ( $$gt[1] eq $ref ) { return ($$gt[0],'0/1'); }
75 return ("$$gt[0],$$gt[1]",'1/2');
87 elsif ( $cons=~/^\+/ ) { return "I$'"; }
88 elsif ( $cons eq '*' ) { return undef; }
89 error("FIXME: could not parse [$cons]\n");
93 # An example of the pileup format:
94 # 1 3000011 C C 32 0 98 1 ^~, A
95 # 1 3002155 * +T/+T 53 119 52 5 +T * 4 1 0
96 # 1 3003094 * -TT/-TT 31 164 60 11 -TT * 5 6 0
97 # 1 3073986 * */-AAAAAAAAAAAAAA 3 3 45 9 * -AAAAAAAAAAAAAA 7 2 0
103 my $fh_in = $$opts{fh_in};
104 my $fh_out = $$opts{fh_out};
105 my ($prev_chr,$prev_pos,$prev_ref);
107 my $ignore_indels = $$opts{snps_only} ? 1 : 0;
108 my $ignore_snps = $$opts{indels_only} ? 1 : 0;
109 my $title = exists($$opts{title}) ? $$opts{title} : 'data';
112 qq[##fileformat=VCFv3.3\n],
113 qq[##INFO=DP,1,Integer,"Total Depth"\n],
114 qq[##FORMAT=GT,1,String,"Genotype"\n],
115 qq[##FORMAT=GQ,1,Integer,"Genotype Quality"\n],
116 qq[##FORMAT=DP,1,Integer,"Read Depth"\n],
117 qq[#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\t$title\n]
120 while (my $line=<$fh_in>)
123 my (@items) = split(/\t/,$line);
124 if ( scalar @items<8 )
126 error("\nToo few columns, does not look like output of 'samtools pileup -c': $line\n");
128 my ($chr,$pos,$ref,$cons,$cons_qual,$snp_qual,$rms_qual,$depth,$a1,$a2) = @items;
133 # An indel is involved.
134 if ( $ignore_indels ) { next; }
136 if (!defined $prev_chr || $chr ne $prev_chr || $pos ne $prev_pos)
138 if ( !$$opts{refseq} ) { error("Cannot do indels without the reference.\n"); }
139 if ( !$refseq ) { $refseq = Fasta->new(file=>$$opts{refseq}); }
140 $ref = $refseq->get_base($chr,$pos);
142 else { $ref = $prev_ref; }
144 # One of the alleles can be a reference and it can come in arbitrary order. In some
145 # cases */* can be encountered. In such a case, look in the additional columns.
146 my ($al1,$al2) = split(m{/},$cons);
147 if ( $al1 eq $al2 && $al1 eq '*' ) { $al1=$a1; $al2=$a2; }
148 my $alt1 = parse_indel($al1);
149 my $alt2 = parse_indel($al2);
150 if ( !$alt1 && !$alt2 ) { error("FIXME: could not parse indel:\n", $line); }
161 elsif ( $alt1 eq $alt2 )
174 if ( $ignore_snps ) { next; }
177 ($alt,$gt) = iupac_to_gtype($ref,$cons);
180 print $fh_out "$chr\t$pos\t.\t$ref\t$alt\t$snp_qual\t0\tDP=$depth\tGT:GQ:DP\t$gt:$cons_qual:$depth\n";
189 #------------- Fasta --------------------
191 # Uses samtools to get a requested base from a fasta file. For efficiency, preloads
192 # a chunk to memory. The size of the cached sequence can be controlled by the 'size'
203 my ($class,@args) = @_;
205 bless $self, ref($class) || $class;
206 if ( !$$self{file} ) { $self->throw(qq[Missing the parameter "file"\n]); }
208 $$self{from} = undef;
210 if ( !$$self{size} ) { $$self{size}=10_000_000; }
211 bless $self, ref($class) || $class;
217 my ($self,$chr,$pos) = @_;
218 my $to = $pos + $$self{size};
219 my $cmd = "samtools faidx $$self{file} $chr:$pos-$to";
221 if ( $? ) { $self->throw("$cmd: $!"); }
222 my $line = shift(@out);
223 if ( !($line=~/^>$chr:(\d+)-(\d+)/) ) { $self->throw("Could not parse: $line"); }
228 while ($line=shift(@out))
233 $$self{chunk} = $chunk;
239 my ($self,$chr,$pos) = @_;
240 if ( !$$self{chr} || $chr ne $$self{chr} || $pos<$$self{from} || $pos>$$self{to} )
242 $self->read_chunk($chr,$pos);
244 my $idx = $pos - $$self{from};
245 return substr($$self{chunk},$idx,1);
250 my ($self,@msg) = @_;