def constructColName2IndexFromHeader(self): """ 2012.8.23 """ self.header = self.next() self.col_name2index = utils.getColName2IndexFromHeader(self.header) return self.col_name2index
def run(self):
"""
"""
if self.debug:
import pdb
pdb.set_trace()
session = self.db_vervet.session
session.begin()
isq_id2data ={}
no_of_total_lines = 0
no_of_isqf_lines = 0
no_of_isqf_in_db = 0
for inputFname in self.inputFnameLs:
reader = csv.reader(open(inputFname), delimiter=figureOutDelimiter(inputFname))
header = reader.next()
colName2Index = utils.getColName2IndexFromHeader(header, skipEmptyColumn=True)
isq_id_index = colName2Index.get('isq_id')
isqf_id_index = colName2Index.get('isqf_id')
read_count_index = colName2Index.get("read_count")
base_count_index = colName2Index.get("base_count")
for row in reader:
isq_id = int(row[isq_id_index])
isqf_id = row[isqf_id_index]
read_count = int(row[read_count_index])
base_count = int(row[base_count_index])
if isq_id not in isq_id2data:
isq_id2data[isq_id] = PassingData(read_count=0, base_count=0)
isq_id2data[isq_id].read_count += read_count
isq_id2data[isq_id].base_count += base_count
if isqf_id and isqf_id!='0':
isqf_id = int(isqf_id)
no_of_isqf_lines += 1
no_of_isqf_in_db += self.updateIndividualSequenceFileReadBaseCount(self.db_vervet, isqf_id=isqf_id, \
read_count=read_count, base_count=base_count)
no_of_total_lines += 1
del reader
logMsg1="%s isqf out of %s were put into db. %s lines in total.\n"%(no_of_isqf_in_db, no_of_isqf_lines, no_of_total_lines)
sys.stderr.write(logMsg1)
counter = 0
real_counter = 0
for isq_id, data in isq_id2data.iteritems():
real_counter += self.updateIndividualSequenceReadBaseCount(self.db_vervet, isq_id=isq_id, \
read_count=data.read_count, base_count=data.base_count, genomeSize=self.genomeSize)
counter += 1
logMsg2="%s isq out of %s were put into db.\n"%(real_counter, counter)
sys.stderr.write(logMsg2)
if self.logFilename:
logF = open(self.logFilename, 'w')
logF.write(logMsg1)
logF.write(logMsg2)
del logF
if self.commit:
self.db_vervet.session.flush()
self.db_vervet.session.commit()
def run(self):
"""
2012.4.3
each input has this as its header:
['alignmentID', 'total_no_of_reads', 'perc_reads_mapped', 'perc_duplicates', 'perc_paired', 'perc_properly_paired', \
'perc_both_mates_mapped', 'perc_singletons',\
'perc_mapped_to_diff_chrs']
"""
if self.debug:
import pdb
pdb.set_trace()
session = self.db_vervet.session
session.begin()
no_of_total_lines = 0
for inputFname in self.inputFnameLs:
reader = csv.reader(open(inputFname), delimiter=figureOutDelimiter(inputFname))
header = reader.next()
colName2Index = utils.getColName2IndexFromHeader(header, skipEmptyColumn=True)
alignment_id_index = colName2Index.get('alignmentID')
total_no_of_reads_index = colName2Index.get('total_no_of_reads')
perc_reads_mapped_index = colName2Index.get("perc_reads_mapped")
perc_duplicates_index = colName2Index.get("perc_duplicates")
perc_paired_index = colName2Index.get("perc_paired")
perc_properly_paired_index = colName2Index.get("perc_properly_paired")
perc_both_mates_mapped_index = colName2Index.get("perc_both_mates_mapped")
perc_singletons_index = colName2Index.get("perc_singletons")
perc_mapped_to_diff_chrs_index = colName2Index.get("perc_mapped_to_diff_chrs")
perc_mapq5_mapped_to_diff_chrs_index = colName2Index.get("perc_mapq5_mapped_to_diff_chrs")
for row in reader:
alignmentID = int(row[alignment_id_index])
alignment = VervetDB.IndividualAlignment.get(alignmentID)
alignment.perc_reads_mapped = float(row[perc_reads_mapped_index])
alignment.perc_duplicates = float(row[perc_duplicates_index])
alignment.perc_paired = float(row[perc_paired_index])
alignment.perc_properly_paired = float(row[perc_properly_paired_index])
alignment.perc_both_mates_mapped = float(row[perc_both_mates_mapped_index])
alignment.perc_singletons = float(row[perc_singletons_index])
alignment.perc_mapped_to_diff_chrs = float(row[perc_mapped_to_diff_chrs_index])
alignment.perc_mapq5_mapped_to_diff_chrs = float(row[perc_mapq5_mapped_to_diff_chrs_index])
alignment.total_no_of_reads = int(float(row[total_no_of_reads_index]))
session.add(alignment)
no_of_total_lines += 1
del reader
sys.stderr.write("%s alignments in total.\n"%(no_of_total_lines))
if self.logFilename:
logF = open(self.logFilename, 'w')
logF.write("%s alignments in total.\n"%(no_of_total_lines))
del logF
if self.commit:
self.db_vervet.session.flush()
self.db_vervet.session.commit()
def _parseHeader(self):
"""
2013.07.17 bugfix, do not reset self.sample_id_ls in the beginning
2012.3.28
add all header content into self.metaInfoLs
except the last header line, which goes into self.sampleIDHeader
2011-11-2
this function is run inside __init__()
"""
self.metaInfoLs = [] #2012.3.28 anything before the "#CHROM" line. each entry is a raw line content, including '\n'
self.sampleIDHeader = [] #2012.3.20 a list of column headers (#CHROM)
self.sample_id2index['ref'] = 0 #ref is at column 0. "ref" must not be equal to any read_group.
self.sample_id_ls.append('ref')
"""
writer = csv.writer(open(outputFname, 'w'), delimiter='\t')
header = ['sample', 'snp_id', 'chr', 'pos', 'qual', 'DP', 'minDP4', 'DP4_ratio', 'MQ']
moreHeader = ['GQ', 'GL', 'SB', 'QD', 'sndHighestGL', 'deltaGL']
#['AF', 'AC','AN', 'Dels', 'HRun', 'HaplotypeScore','MQ0', 'QD'] #2011-3-4 useless
if VCFOutputType==2:
header += moreHeader
chr_pure_number_pattern = re.compile(r'[a-z_A-Z]+(\d+)')
chr_number_pattern = re.compile(r'chr(\d+)')
"""
counter = 0
real_counter = 0
for line in self.inf:
if line[:6]=='#CHROM':
line = line.strip() #get rid of the trailing \n
row = line.split('\t')
self.sampleIDHeader = row[self.sampleStartingColumn:]
self.header = row[:]
self.headerWithoutHash= row[:]
self.headerWithoutHash[0] = 'CHROM' #discard the #
self.col_name2index = getColName2IndexFromHeader(self.headerWithoutHash, skipEmptyColumn=True)
self.col_index_individual_name_ls = self._getIndividual2ColIndex(self.headerWithoutHash, self.col_name2index)
for individual_col_index, individual_name in self.col_index_individual_name_ls:
read_group = individual_name.strip()
if read_group not in self.sample_id2index:
self.sample_id2index[read_group] = len(self.sample_id2index)
self.sample_id_ls.append(read_group)
break # "#CHROM" is the last line of the self.headerWithoutHash
elif line[0]=='#': #2011-3-4
self.metaInfoLs.append(line)
#continue
else: #leave everything for parseFile or parseIter
break
def processHeader(self, reader=None, extendHeader=None, chrLengthHeader = 'chrLength'):
"""
2012.8.7
modularize so that AddHetFractionToVCFtoolsHWE could inherit
"""
header = reader.next()
self.originalHeader = header
self.col_name2index = utils.getColName2IndexFromHeader(self.originalHeader, skipEmptyColumn=True)
self.originalHeaderLength = len(header)
header.extend(extendHeader)
if self.divideByLength:
i = self.divideStartingColumn
while (i<self.originalHeaderLength):
statColumnHeader = header[i]
header.append("%s_div_by_%s"%(statColumnHeader, chrLengthHeader))
i += 1;
return header
def getBamBaseFname2MonkeyID_WUSTLDNAData(self, inputFname, ):
"""
2011-8-3
from WUSTL
the input looks like this:
# FlowCell Lane Index Sequence Library Common Name Bam Path MD5
1 64J6AAAXX 1 VCAC-2007002-1-lib1 African Green Monkey /gscmnt/sata755/production/csf_111215677/gerald_64J6AAAXX_1.bam /gscmnt/sata755/production/csf_111215677/gerald_64J6AAAXX_1.bam.md5
2 64J6AAAXX 2 VCAC-2007006-1-lib1 African Green Monkey /gscmnt/sata751/production/csf_111215675/gerald_64J6AAAXX_2.bam /gscmnt/sata751/production/csf_111215675/gerald_64J6AAAXX_2.bam.md5
"""
sys.stderr.write("Getting bamBaseFname2MonkeyID dictionary ...")
bamBaseFname2MonkeyID = {}
reader = csv.reader(open(inputFname), delimiter='\t')
header = reader.next()
col_name2index = getColName2IndexFromHeader(header, skipEmptyColumn=True)
monkeyIDIndex = col_name2index.get("Library")
if monkeyIDIndex is None: #2012.6.7
monkeyIDIndex = col_name2index.get("library")
bamFnameIndex = col_name2index.get("Bam Path")
if bamFnameIndex is None: #2012.2.9
bamFnameIndex = col_name2index.get("BAM Path")
if bamFnameIndex is None: #2012.2.9
bamFnameIndex = col_name2index.get("BAM")
if bamFnameIndex is None: #2012.6.7
bamFnameIndex = col_name2index.get("bam pathway")
#monkeyIDPattern = re.compile(r'\w+-(\w+)-\d+-\w+') # i.e. VCAC-2007002-1-lib1
monkeyIDPattern = re.compile(r'\w+-(\w+)-\w+-\w+') # 2012.5.29 i.e. VCAC-VGA00006-AGM0075-lib1 ,
# VCAC-VZC1014-AGM0055-lib1, VCAC-1996031-VRV0265-lib2a, VCAC-VKD7-361-VKD7-361-lib1 (VKD7 is to be taken),
for row in reader:
monkeyID = row[monkeyIDIndex]
pa_search = monkeyIDPattern.search(monkeyID)
if pa_search:
monkeyID = pa_search.group(1)
else:
sys.stderr.write("Warning: could not parse monkey ID from %s. Ignore.\n"%(monkeyID))
continue
bamFname = row[bamFnameIndex]
bamBaseFname = os.path.split(bamFname)[1]
bamBaseFname2MonkeyID[bamBaseFname] = monkeyID
sys.stderr.write("%s entries.\n"%(len(bamBaseFname2MonkeyID)))
return bamBaseFname2MonkeyID
def smartReadHeader(self, headerPattern=None, commentPattern=None): """ Note: If an input file does not have a header, this function over-reads by one line (stored in self._row) so need to process the last self._row before further reading 2013.08.30 read the header, while ignoring lines fitting the comment pattern and construct col_name2index when a line matching headerPattern is encountered """ if headerPattern is None: headerPattern = self.headerPattern if commentPattern is None: commentPattern = self.commentPattern row = self.next() while commentPattern.search(row[0]): #passing all comments self.comment_row_list.append(row) row = self.next() if headerPattern.search(row[0]): self.header = row self.col_name2index = utils.getColName2IndexFromHeader(self.header) else: self.col_name2index = None return self.col_name2index
def discoverFromVCF(cls, inputFname, outputFname, refFastaFname=None, VCFOutputType=2, \
minMinorAlleleCoverage=1/4., maxMinorAlleleCoverage=3/4.,\
maxNoOfReads=2., minNoOfReads=1/4., \
maxNoOfReadsForGenotypingError=1, maxMajorAlleleCoverage=7/8., maxNoOfReadsForAllSamples=1000,\
nt_set = set(['a','c','g','t','A','C','G','T']), isqID2coverage=None, defaultCoverage=10, \
outputDelimiter='\t',\
report=0, site_type=1):
"""
2011-9-2
add argument isqID2coverage, defaultCoverage
2011-8-26
add argument site_type
function is also more robust against missing fields etc.
2011-7-20
copied from discoverHetsFromVCF() of vervet.src.misc
2011-3-24
add maxMinorAlleleCoverage
Even a heterozygote's MAC is within [minMinorAlleleCoverage, maxMinorAlleleCoverage], it could still be
a homozygous SNP.
2011-3-4
VCF output by GATK has a different format
argument VCFOutputType
1: output by samtools's vcfutils.pl
2: output by GATK
2011-1-6
inputFname is VCF output by "vcfutils.pl varFilter" of samtools
"""
import csv
from pymodule.utils import runLocalCommand, getColName2IndexFromHeader
sys.stderr.write("Looking for heterozygous SNPs in %s (%s<=MAC<=%s).\n"%(os.path.basename(inputFname), \
minMinorAlleleCoverage, maxMinorAlleleCoverage))
reader =csv.reader(open(inputFname), delimiter='\t')
read_group2col_index = {'ref':0} #ref is at column 0. "ref" must not be equal to any read_group.
read_group2coverage = {} #2011-9-2
locus_id2row_index = {}
data_matrix = []
tid2refName = {} #dictionary storing the target references which have SNP calls
refNameSet = set()
"""
writer = csv.writer(open(outputFname, 'w'), delimiter='\t')
header = ['sample', 'snp_id', 'chr', 'pos', 'qual', 'DP', 'minDP4', 'DP4_ratio', 'MQ']
moreHeader = ['GQ', 'GL', 'SB', 'QD', 'sndHighestGL', 'deltaGL']
#['AF', 'AC','AN', 'Dels', 'HRun', 'HaplotypeScore','MQ0', 'QD'] #2011-3-4 useless
if VCFOutputType==2:
header += moreHeader
chr_pure_number_pattern = re.compile(r'[a-z_A-Z]+(\d+)')
chr_number_pattern = re.compile(r'chr(\d+)')
"""
individual_name2col_index = None
col_name2index = None
counter = 0
real_counter = 0
for row in reader:
if row[0] =='#CHROM':
row[0] = 'CHROM' #discard the #
header = row
col_name2index = getColName2IndexFromHeader(header, skipEmptyColumn=True)
individual_name2col_index = cls.getIndividual2ColIndex(header, col_name2index)
continue
elif row[0][0]=='#': #2011-3-4
continue
"""
if chr_number_pattern.search(row[0]):
chr = chr_number_pattern.search(row[0]).group(1)
elif chr_pure_number_pattern.search(row[0]):
chr = chr_pure_number_pattern.search(row[0]).group(1)
else:
sys.stderr.write("Couldn't parse the chromosome number/character from %s.\n Exit.\n"%(row[0]))
sys.exit(4)
"""
chr = row[0]
refNameSet.add(chr)
pos = row[1]
quality = row[5]
outputHet= False
info = row[7]
info_ls = info.split(';')
info_tag2value = {}
for info in info_ls:
try:
tag, value = info.split('=')
except:
#sys.stderr.write("Error in splitting %s by =.\n"%info) ###Error in splitting DS by =.
continue
info_tag2value[tag] = value
current_locus = '%s_%s'%(chr, pos)
refBase = row[col_name2index['REF']]
altBase = row[col_name2index['ALT']]
if VCFOutputType==2: #2011-3-4 GATK
format_column = row[col_name2index['FORMAT']]
format_column_ls = format_column.split(':')
format_column_name2index = getColName2IndexFromHeader(format_column_ls)
data_row = ['NA']*(len(individual_name2col_index)+1) # extra 1 for the ref
allele2count = {}
for individual_name, individual_col_index in individual_name2col_index.iteritems():
read_group = individual_name
if read_group not in read_group2col_index:
read_group2col_index[read_group] = len(read_group2col_index)
#2011-9-2
if isqID2coverage:
try:
isqID = read_group.split('_')[1]
isqID = int(isqID)
coverage = isqID2coverage.get(isqID, defaultCoverage)
except:
sys.stderr.write('Except type: %s\n'%repr(sys.exc_info()))
import traceback
traceback.print_exc()
sys.stderr.write("Coverage for %s not available. use default=%s.\n"%(read_group, defaultCoverage))
coverage = defaultCoverage
else:
coverage = defaultCoverage
read_group2coverage[read_group] = coverage
coverage = read_group2coverage[read_group]
genotype_data = row[individual_col_index]
genotype_data_ls = genotype_data.split(':')
genotype_call_index = format_column_name2index.get('GT')
genotype_quality_index = format_column_name2index.get('GQ')
if genotype_quality_index is None:
genotype_quality_index = format_column_name2index.get('DP')
depth_index = format_column_name2index.get("DP")
#GL_index = format_column_name2index.get('GL')
if len(genotype_data_ls)<len(format_column_name2index):
continue
if depth_index is None or genotype_call_index is None:
continue
#genotype_quality = genotype_data_ls[genotype_quality_index]
genotype_call = genotype_data_ls[genotype_call_index]
depth = int(genotype_data_ls[depth_index])
if depth>maxNoOfReads*coverage or depth<minNoOfReads*coverage: #2011-3-29 skip. coverage too high or too low
continue
allele = 'NA'
if genotype_call=='0/1' or genotype_call =='1/0': #heterozygous, the latter notation is never used though.
"""
GL_list = genotype_data_ls[GL_index]
GL_list = GL_list.split(',')
GL_list = map(float, GL_list)
GL = GL_list[1]
sndHighestGL = max([GL_list[0], GL_list[2]])
deltaGL = GL-sndHighestGL
"""
AD = genotype_data_ls[format_column_name2index.get('AD')]
AD = map(int, AD.split(','))
minorAlleleCoverage = min(AD)
majorAlleleCoverage = max(AD)
if minorAlleleCoverage<=maxMinorAlleleCoverage*coverage and minorAlleleCoverage>=minMinorAlleleCoverage*coverage \
and majorAlleleCoverage<=maxMajorAlleleCoverage*coverage:
DP4_ratio = float(AD[0])/AD[1]
allele = '%s%s'%(refBase, altBase)
"""
data_row = [individual_name, 'chr%s:%s'%(chr, pos), chr, pos, quality, \
depth, minorAlleleCoverage, DP4_ratio,\
info_tag2value.get('MQ'), genotype_quality, GL,\
info_tag2value.get('SB'), info_tag2value.get('QD'), sndHighestGL, deltaGL]
#for i in range(3, len(moreHeader)):
# info_tag = moreHeader[i]
# data_row.append(info_tag2value.get(info_tag))
writer.writerow(data_row)
"""
elif genotype_call=='./.': #missing
continue
elif genotype_call =='1/1':
allele = '%s%s'%(altBase, altBase)
elif genotype_call =='0/0':
allele = '%s%s'%(refBase, refBase)
col_index = read_group2col_index.get(read_group)
data_row[col_index] = allele
if allele!='NA':
if allele not in allele2count:
allele2count[allele] = 0
allele2count[allele] += 1
if len(allele2count)>site_type-1: #whether polymorphic across samples or all sites in vcf
real_counter += 1
locus_id2row_index[current_locus] = len(locus_id2row_index)
data_matrix.append(data_row)
"""
elif VCFOutputType==1: #samtools. 2011-7-20 outdated.
sample_id = row[8]
for tag in info_tag2value.keys():
value = info_tag2value.get(tag)
if tag=='DP4':
tag = 'DP4_ratio'
value = value.split(',')
value = map(int, value)
no_of_ref_allele = sum(value[0:2])
no_of_non_ref_allele = sum(value[2:])
MAC = min(no_of_ref_allele, no_of_non_ref_allele)
if MAC<=maxMinorAlleleCoverage and MAC>=minMinorAlleleCoverage:
outputHet = True
value = float(no_of_ref_allele)/no_of_non_ref_allele
info_tag2value['minDP4'] = min(no_of_ref_allele, no_of_non_ref_allele)
else:
value = None
info_tag2value[tag] = value
if outputHet:
real_counter += 1
output_row = [sample_id, 'chr%s:%s'%(chr, pos), chr, pos, quality, info_tag2value.get('DP'), \
info_tag2value.get('minDP4'), info_tag2value.get('DP4_ratio'), info_tag2value.get('MQ')]
writer.writerow(output_row)
"""
counter += 1
if counter%2000==0 and report:
sys.stderr.write("%s\t%s\t%s"%("\x08"*80, counter, real_counter))
del reader
cls.outputCallMatrix(data_matrix, refFastaFname, outputFname=outputFname, refNameSet=refNameSet, \
read_group2col_index=read_group2col_index, \
locus_id2row_index=locus_id2row_index, outputDelimiter=outputDelimiter)
sys.stderr.write("%s\t%s\t%s.\n"%("\x08"*80, counter, real_counter))
def run(self):
"""
2012.5.7 new input looks like this (tab-delimited):
alignmentID total_base_count sampled_base_count meanDepth medianDepth modeDepth
100 1005506 301614 70.0441756682 9.0 8.0
27 1005506 301614 70.0441756682 9.0 8.0
2012.4.3
each input looks like this:
sample_id total mean granular_third_quartile granular_median granular_first_quartile %_bases_above_15
553_2_VRC_ref_GA_vs_524 2434923137 8.25 11 9 6 4.4
Total 2434923137 8.25 N/A N/A N/A
554_3_Barbados_GA_vs_524 2136011136 7.23 11 8 6 3.5
Total 2136011136 7.23 N/A N/A N/A
...
"""
if self.debug:
import pdb
pdb.set_trace()
session = self.db_vervet.session
session.begin()
no_of_total_lines = 0
for inputFname in self.inputFnameLs:
reader = csv.reader(open(inputFname), delimiter=figureOutDelimiter(inputFname))
header = reader.next()
col_name2index = utils.getColName2IndexFromHeader(header, skipEmptyColumn=True)
sample_id_index = col_name2index.get("alignmentID")
total_base_count_index = col_name2index.get('total_base_count')
mean_depth_index = col_name2index.get("meanDepth")
median_depth_index = col_name2index.get("medianDepth")
mode_depth_index = col_name2index.get("modeDepth")
for row in reader:
sample_id = row[sample_id_index]
if sample_id=='Total': #ignore rows with this as sample id
continue
alignment_id = int(sample_id.split("_")[0])
total_base_count = int(row[total_base_count_index])
mean_depth = float(row[mean_depth_index])
median_depth = float(row[median_depth_index])
mode_depth = float(row[mode_depth_index])
individual_alignment = VervetDB.IndividualAlignment.get(alignment_id)
individual_alignment.pass_qc_read_base_count = total_base_count #2012.9.17 no longer trustworthy because CalculateMedianMeanOfInputColumn skips data.
individual_alignment.mean_depth = mean_depth
individual_alignment.median_depth = median_depth
individual_alignment.mode_depth = mode_depth
session.add(individual_alignment)
no_of_total_lines += 1
del reader
sys.stderr.write("%s alignments in total.\n"%(no_of_total_lines))
if self.logFilename:
logF = open(self.logFilename, 'w')
logF.write("%s alignments in total.\n"%(no_of_total_lines))
del logF
if self.commit:
self.db_vervet.session.flush()
self.db_vervet.session.commit()
def parseOneVCFRow(row, col_name2index, col_index_individual_name_ls, sample_id2index, minDepth=1,\
dataEntryType=1):
"""
2014.01.08 fix a bug that skips calls and shortens data_row.
2012.9.6 turn pos into integer
2012.5.10
complete representation of one locus
2012.1.17
common snippet split out of VCFFile & VCFRecord
row is a list of input columns from one VCF file line
dataEntryType
1: each cell is base call
2: each cell is a dictionary {'GT': base-call, 'DP': depth}
"""
chromosome = row[0]
pos = int(row[1]) #2012.9.6 turn pos into integer
vcf_locus_id=row[2]
quality = row[5]
filter=row[6]
info = row[7]
format = row[8]
info_ls = info.split(';')
info_tag2value = {}
for info_entry in info_ls:
try:
tag, value = info_entry.split('=')
except:
#sys.stderr.write("Error in splitting %s by =.\n"%info) ###Error in splitting DS by =.
continue
info_tag2value[tag] = value
locus_id = (chromosome, pos)
refBase = row[col_name2index['REF']]
altBase = row[col_name2index['ALT']]
altBaseLs = altBase.split(',') #altBase could be just "C" or "C,G" (multi-nucleotide)
alleleLs = [refBase] + altBaseLs
alleleNumber2Base = {'.':'NA'}
for i in xrange(len(alleleLs)):
alleleNumber2Base[repr(i)] = alleleLs[i]
format_column = row[col_name2index['FORMAT']]
format_column_ls = format_column.split(':')
format_column_name2index = getColName2IndexFromHeader(format_column_ls)
if dataEntryType==1:
data_row = ['NA']*(len(col_index_individual_name_ls)+1) # extra 1 for the ref
data_row[0] = refBase
else:
data_row = [None]*(len(col_index_individual_name_ls)+1) # extra 1 for the ref
data_row[0] = {'GT':refBase, 'DP':-1}
genotypeCall2Count = {}
for individual_col_index, individual_name in col_index_individual_name_ls:
individual_name = individual_name
if individual_name not in sample_id2index:
sample_id2index[individual_name] = len(sample_id2index)
#coverage = read_group2coverage[individual_name]
genotype_data = row[individual_col_index]
genotype_data_ls = genotype_data.split(':')
genotype_call_index = format_column_name2index.get('GT')
genotype_quality_index = format_column_name2index.get('GQ')
if genotype_quality_index is None:
genotype_quality_index = format_column_name2index.get('DP')
depth_index = format_column_name2index.get("DP")
#GL_index = format_column_name2index.get('GL')
genotypeCallInBase = 'NA'
if genotype_call_index is not None and len(genotype_data_ls)>0:
# or (genotype_call_index is not None and len(genotype_data_ls)<=genotype_call_index): #<len(format_column_name2index): #this genotype call is probably empty "./." due to no reads
#genotype_quality = genotype_data_ls[genotype_quality_index]
if genotype_call_index is not None and len(genotype_data_ls)>genotype_call_index:
genotype_call = genotype_data_ls[genotype_call_index]
else:
genotype_call = './.' #missing
callData = {}
if genotype_call!='./.' and genotype_call!='.' and genotype_call!='.|.': #missing data
patternSearchResult = diploidGenotypePattern.search(genotype_call)
if patternSearchResult:
allele1 = alleleNumber2Base[patternSearchResult.group(1)]
allele2 = alleleNumber2Base[patternSearchResult.group(2)]
if allele1!='N' and allele2!='N':
genotypeCallInBase = '%s%s'%(allele1, allele2)
if depth_index is not None:
if len(genotype_data_ls)>depth_index:
depth = genotype_data_ls[depth_index]
else:
depth = '.' #missing DP
if depth=='.': #this means depth=0
depth = 0
else:
depth = int(depth)
if minDepth>0 and depth<minDepth: #no read. samtools would still assign ref/ref to this individual
genotypeCallInBase = 'NA' #set it to missing
#if depth>maxNoOfReads*coverage or depth<minNoOfReads*coverage: #2011-3-29 skip. coverage too high or too low
# continue
callData['DP'] = depth
"""
if genotype_call=='0/1' or genotype_call =='1/0': #heterozygous, the latter notation is never used though.
allele = '%s%s'%(refBase, altBase)
GL_list = genotype_data_ls[GL_index]
GL_list = GL_list.split(',')
GL_list = map(float, GL_list)
GL = GL_list[1]
sndHighestGL = max([GL_list[0], GL_list[2]])
deltaGL = GL-sndHighestGL
AD = genotype_data_ls[format_column_name2index.get('AD')]
AD = map(int, AD.split(','))
minorAlleleCoverage = min(AD)
majorAlleleCoverage = max(AD)
if minorAlleleCoverage<=minorAlleleDepthUpperBoundCoeff*coverage and \
minorAlleleCoverage>=minorAlleleDepthLowerBoundCoeff*coverage and \
majorAlleleCoverage<=majorAlleleDepthUpperBoundCoeff*coverage:
DP4_ratio = float(AD[0])/AD[1]
allele = '%s%s'%(refBase, altBase)
elif genotype_call=='./.' or genotype_call=='.|.': #missing
allele = 'NA'
elif genotype_call =='1/1' or genotype_call =='1|1':
allele = '%s%s'%(altBase, altBase)
elif genotype_call =='0/0' or genotype_call=='0|0':
allele = '%s%s'%(refBase, refBase)
"""
col_index = sample_id2index.get(individual_name)
if dataEntryType==1:
data_row[col_index] = genotypeCallInBase
else:
callData['GT'] = genotypeCallInBase
data_row[col_index] = callData
if genotypeCallInBase!='NA':
if genotypeCallInBase not in genotypeCall2Count:
genotypeCall2Count[genotypeCallInBase] = 0
genotypeCall2Count[genotypeCallInBase] += 1
return PassingData(chr=chromosome, chromosome=chromosome, pos=pos, position=pos, locus_id=locus_id, quality=quality, \
info_tag2value=info_tag2value, \
refBase=refBase, altBase=altBase, \
alleleLs=alleleLs, alleleNumber2Base=alleleNumber2Base, genotypeCall2Count=genotypeCall2Count, data_row=data_row,\
info=info, format=format, filter=filter, vcf_locus_id=vcf_locus_id, \
format_column_name2index=format_column_name2index, format_column_ls=format_column_ls)
