3 # VCF specs: http://www.1000genomes.org/wiki/doku.php?id=1000_genomes:analysis:vcfv3.2
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 " -r, -refseq <file.fa> The reference sequence, required when indels are present.\n",
27 " -h, -?, --help This help message.\n",
36 $opts{fh_in} = *STDIN;
37 $opts{fh_out} = *STDOUT;
39 while (my $arg=shift(@ARGV))
41 if ( $arg eq '-r' || $arg eq '--refseq' ) { $opts{refseq}=shift(@ARGV); next; }
42 if ( $arg eq '-?' || $arg eq '-h' || $arg eq '--help' ) { error(); }
44 error("Unknown parameter \"$arg\". Run -h for help.\n");
60 if ( !exists($iupac{$base}) )
62 if ( $ref eq $base ) { return ('.','0|0'); }
65 my $gt = $iupac{$base};
66 if ( $$gt[0] eq $ref ) { return ($$gt[1],'0|1'); }
67 elsif ( $$gt[1] eq $ref ) { return ($$gt[0],'0|1'); }
68 return ("$$gt[0],$$gt[1]",'1|2');
80 elsif ( $cons=~/^\+/ ) { return "I$'"; }
81 elsif ( $cons eq '*' ) { return undef; }
82 error("FIXME: could not parse [$cons]\n");
86 # An example of the pileup format:
87 # 1 3000011 C C 32 0 98 1 ^~, A
88 # 1 3002155 * +T/+T 53 119 52 5 +T * 4 1 0
89 # 1 3003094 * -TT/-TT 31 164 60 11 -TT * 5 6 0
90 # 1 3073986 * */-AAAAAAAAAAAAAA 3 3 45 9 * -AAAAAAAAAAAAAA 7 2 0
96 my $fh_in = $$opts{fh_in};
97 my $fh_out = $$opts{fh_out};
98 my ($prev_chr,$prev_pos,$prev_ref);
101 while (my $line=<$fh_in>)
104 my ($chr,$pos,$ref,$cons,$cons_qual,$snp_qual,$rms_qual,$depth,@items) = split(/\t/,$line);
109 # An indel is involved.
110 if ($chr ne $prev_chr || $pos ne $prev_pos)
112 if ( !$$opts{refseq} ) { error("Cannot do indels without the reference.\n"); }
113 if ( !$refseq ) { $refseq = Fasta->new(file=>$$opts{refseq}); }
114 $ref = $refseq->get_base($chr,$pos);
116 else { $ref = $prev_ref; }
118 # One of the alleles can be a reference and it can come in arbitrary order
119 my ($al1,$al2) = split(m{/},$cons);
120 my $alt1 = parse_indel($al1);
121 my $alt2 = parse_indel($al2);
122 if ( !$alt1 && !$alt2 ) { error("FIXME: could not parse indel:\n", $line); }
123 if ( $alt1 && $alt2 && $alt1 eq $alt2 ) { $alt2=''; }
143 ($alt,$gt) = iupac_to_gtype($ref,$cons);
146 print $fh_out "$chr\t$pos\t.\t$ref\t$alt\t$snp_qual\t0\t\tGT:GQ:DP\t$gt:$cons_qual:$depth\n";
155 #------------- Fasta --------------------
157 # Uses samtools to get a requested base from a fasta file. For efficiency, preloads
158 # a chunk to memory. The size of the cached sequence can be controlled by the 'size'
169 my ($class,@args) = @_;
170 my $self = @args ? {@args} : {};
171 if ( !$$self{file} ) { $self->throw(qq[Missing the parameter "file"\n]); }
173 $$self{from} = undef;
175 if ( !$$self{size} ) { $$self{size}=10_000_000; }
176 bless $self, ref($class) || $class;
182 my ($self,$chr,$pos) = @_;
183 my $to = $pos + $$self{size};
184 my $cmd = "samtools faidx $$self{file} $chr:$pos-$to";
186 if ( $? ) { $self->throw("$cmd: $!"); }
187 my $line = shift(@out);
188 if ( !($line=~/^>$chr:(\d+)-(\d+)/) ) { $self->throw("Could not parse: $line"); }
193 while ($line=shift(@out))
198 $$self{chunk} = $chunk;
204 my ($self,$chr,$pos) = @_;
205 if ( !$$self{chr} || $chr ne $$self{chr} || $pos<$$self{from} || $pos>$$self{to} )
207 $self->read_chunk($chr,$pos);
209 my $idx = $pos - $$self{from};
210 return substr($$self{chunk},$idx,1);
215 my ($self,@msg) = @_;