1 | #Dan Blankenberg |
---|
2 | #See: http://1000genomes.org/wiki/doku.php?id=1000_genomes:analysis:vcf3.3 |
---|
3 | #See: http://1000genomes.org/wiki/doku.php?id=1000_genomes:analysis:variant_call_format |
---|
4 | |
---|
5 | class VariantCall( object ): |
---|
6 | version = None |
---|
7 | header_startswith = None |
---|
8 | required_header_fields = None |
---|
9 | required_header_length = None |
---|
10 | |
---|
11 | @classmethod |
---|
12 | def get_class_by_format( cls, format ): |
---|
13 | assert format in VCF_FORMATS, 'Unknown format type specified: %s' % format |
---|
14 | return VCF_FORMATS[ format ] |
---|
15 | |
---|
16 | def __init__( self, vcf_line, metadata, sample_names ): |
---|
17 | raise Exception( 'Abstract Method' ) |
---|
18 | |
---|
19 | class VariantCall33( VariantCall ): |
---|
20 | version = 'VCFv3.3' |
---|
21 | header_startswith = '#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO' |
---|
22 | required_header_fields = header_startswith.split( '\t' ) |
---|
23 | required_header_length = len( required_header_fields ) |
---|
24 | |
---|
25 | def __init__( self, vcf_line, metadata, sample_names ): |
---|
26 | # Raw line is needed for indexing file. |
---|
27 | self.raw_line = vcf_line |
---|
28 | self.line = vcf_line.rstrip( '\n\r' ) |
---|
29 | self.metadata = metadata |
---|
30 | self.sample_names = sample_names |
---|
31 | self.format = None |
---|
32 | self.sample_values = [] |
---|
33 | |
---|
34 | #parse line |
---|
35 | self.fields = self.line.split( '\t' ) |
---|
36 | if sample_names: |
---|
37 | assert len( self.fields ) == self.required_header_length + len( sample_names ) + 1, 'Provided VCF line (%s) has wrong length (expected: %i)' % ( self.line, self.required_header_length + len( sample_names ) + 1 ) |
---|
38 | else: |
---|
39 | assert len( self.fields ) == self.required_header_length, 'Provided VCF line (%s) has wrong length (expected: %i)' % ( self.line, self.required_header_length) |
---|
40 | self.chrom, self.pos, self.id, self.ref, self.alt, self.qual, self.filter, self.info = self.fields[ :self.required_header_length ] |
---|
41 | self.pos = int( self.pos ) |
---|
42 | self.alt = self.alt.split( ',' ) |
---|
43 | self.qual = float( self.qual ) |
---|
44 | if len( self.fields ) > self.required_header_length: |
---|
45 | self.format = self.fields[ self.required_header_length ].split( ':' ) |
---|
46 | for sample_value in self.fields[ self.required_header_length + 1: ]: |
---|
47 | self.sample_values.append( sample_value.split( ':' ) ) |
---|
48 | |
---|
49 | class VariantCall40( VariantCall33 ): |
---|
50 | version = 'VCFv4.0' |
---|
51 | def __init__( self, vcf_line, metadata, sample_names ): |
---|
52 | VariantCall33.__init__( self, vcf_line, metadata, sample_names) |
---|
53 | |
---|
54 | #VCF Format version lookup dict |
---|
55 | VCF_FORMATS = {} |
---|
56 | for format in [ VariantCall33, VariantCall40 ]: |
---|
57 | VCF_FORMATS[format.version] = format |
---|
58 | |
---|
59 | class Reader( object ): |
---|
60 | def __init__( self, fh ): |
---|
61 | self.vcf_file = fh |
---|
62 | self.metadata = {} |
---|
63 | self.header_fields = None |
---|
64 | self.metadata_len = 0 |
---|
65 | self.sample_names = [] |
---|
66 | self.vcf_class = None |
---|
67 | |
---|
68 | # Read file metadata. |
---|
69 | while True: |
---|
70 | line = self.vcf_file.readline() |
---|
71 | self.metadata_len += len( line ) |
---|
72 | assert line, 'Invalid VCF file provided.' |
---|
73 | line = line.rstrip( '\r\n' ) |
---|
74 | if self.vcf_class and line.startswith( self.vcf_class.header_startswith ): |
---|
75 | # read the header fields, ignoring any blank tabs, which GATK |
---|
76 | # VCF produces after the sample |
---|
77 | self.header_fields = [l for l in line.split( '\t' ) if l] |
---|
78 | if len( self.header_fields ) > self.vcf_class.required_header_length: |
---|
79 | for sample_name in self.header_fields[ self.vcf_class.required_header_length + 1 : ]: |
---|
80 | self.sample_names.append( sample_name ) |
---|
81 | break |
---|
82 | assert line.startswith( '##' ), 'Non-metadata line found before header' |
---|
83 | line = line[2:] #strip ## |
---|
84 | metadata = line.split( '=', 1 ) |
---|
85 | metadata_name = metadata[ 0 ] |
---|
86 | if len( metadata ) == 2: |
---|
87 | metadata_value = metadata[ 1 ] |
---|
88 | else: |
---|
89 | metadata_value = None |
---|
90 | if metadata_name in self.metadata: |
---|
91 | if not isinstance( self.metadata[ metadata_name ], list ): |
---|
92 | self.metadata[ metadata_name ] = [ self.metadata[ metadata_name ] ] |
---|
93 | self.metadata[ metadata_name ].append( metadata_value ) |
---|
94 | else: |
---|
95 | self.metadata[ metadata_name ] = metadata_value |
---|
96 | if metadata_name == 'fileformat': |
---|
97 | self.vcf_class = VariantCall.get_class_by_format( metadata_value ) |
---|
98 | def next( self ): |
---|
99 | line = self.vcf_file.readline() |
---|
100 | if not line: |
---|
101 | raise StopIteration |
---|
102 | return self.vcf_class( line, self.metadata, self.sample_names ) |
---|
103 | def __iter__( self ): |
---|
104 | while True: |
---|
105 | yield self.next() |
---|