def run(args):
scoreNames,bedRegions = sniffBed(args.bedfile)
if args.names != None:
scoreNames.difference_update(args.names)
regionsToRemove = set()
for i,r in bedRegions:
if r.name not in scoreNames:
regionsToRemove.add(i)
for i in regionsToRemove:
del bedRegions[i]
vcffile = open(args.infile,'r')
outfile = open(args.outfile,'w')
takenTags = set()
for line in vcffile:
if len(line) <= 1:
continue
elif line.startswith('##'):
if line.startswith("##INFO"):
newTag = line[line.find("ID=")+3:]
newTag = newTag[:newTag.find(',')]
takenTags.add(newTag)
outfile.write(line)
elif line.startswith('#'):
for n in scoreNames:
dupCount = 2
newTag = n
while newTag in takenTags:
newTag = n + str(dupCount)
dupCount += 1
takenTags.add(newTag)
outfile.write("##INFO=<ID=%s,Number=.,Type=Float,Description=\"User column added with addBEDtoVCF.py\">\n" % newTag)
outfile.write(line)
else:
line = vcfLine(line.strip().split('\t'))
line.extractChrAndPos()
for b in bedRegions:
if b.contains(line.chromosome, line.position):
line.extractInfo()
line.info[b.name] = str(b.score)
outfile.write(str(line))
vcffile.close()
outfile.close()
def run(args):
print 'Counting values...'
max_strings = args.max_strings
ignoreStringCounts = False
if max_strings <= 0:
ignoreStringCounts = True
max_strings = MAX_INFO_STRINGS
infoFields = extractInfoFields(args.infile, max_strings)
validFields = set()
for k,f in infoFields.iteritems():
if not ignoreStringCounts and f.maxedOut:
continue
if args.preserve_info != None and k not in args.preserve_info:
continue
if args.remove_info != None and k in args.remove_info:
continue
validFields.add(k)
print 'Writing file...'
outfile = open(args.outfile, 'w')
infile = open(args.infile, 'r')
for line in infile:
line = line
if len(line) <= 1:
continue
elif line.startswith("##INFO"):
newTag = line[line.find("ID=")+3:]
newTag = newTag[:newTag.find(',')]
if not infoFields.has_key(newTag):
raise Exception("Second pass lost info tag:\t%s" % newTag)
if newTag in validFields:
outfile.write(line)
elif line.startswith("#"):
outfile.write(line)
else:
line = vcfLine(line.strip().split('\t'))
line.extractInfo()
keys = line.info.keys()
for k in keys:
if k not in validFields:
del line.info[k]
outfile.write(str(line))
infile.close()
outfile.close()
def extractInfoFields(path,max_strings=MAX_INFO_STRINGS,tickFunction=None,numTicks=1000):
infoFields = {}
tickInterval = os.path.getsize(path)/numTicks
nextTick = 0
if path.endswith('gz'):
infile = gzip.open(path,'rb')
else:
infile = open(path,'rb')
for line in infile:
if tickFunction != None:
if infile.tell() > nextTick:
if not tickFunction():
infile.close()
return None
nextTick += tickInterval
line = line.strip()
if len(line) <= 1:
continue
elif line.startswith("##INFO"):
newTag = line[line.find("ID=")+3:]
newTag = newTag[:newTag.find(',')]
if infoFields.has_key(newTag):
raise Exception("Duplicate INFO ID or use of reserved ID:\t%s" % newTag)
infoFields[newTag] = infoDetails(newTag, max_strings, False)
elif line.startswith("#"):
continue
else:
line = vcfLine(line.split('\t'))
line.extractInfo()
for k,v in line.info.iteritems():
if not infoFields.has_key(k):
raise Exception("Missing ##INFO pragma for: %s" % k)
else:
infoFields[k].addArbitraryValue(v)
infile.close()
return infoFields
def __init__(self, line):
if len(line.strip()) == 0:
self.type = vcfKey.EMPTY
elif line.startswith('##'):
line = line.lower()
if line.startswith('##fileformat'):
self.type = vcfKey.FIRSTLINE
elif line.startswith('##contig'):
self.type = vcfKey.CONTIG
contigID = line[line.find('ID=') + 3:]
contigID = contigID[:contigID.find(',')]
self.chromosome = standardizeChromosome(contigID)
else:
self.type = vcfKey.OTHER_META
self.line = line
elif line.startswith('#'):
self.type = vcfKey.HEADER
else:
self.type = vcfKey.REGULAR
temp = vcfLine(line.strip().split('\t'))
temp.extractChrAndPos()
self.chromosome = temp.chromosome
self.position = temp.position
def __init__(self, line):
if len(line.strip()) == 0:
self.type = vcfKey.EMPTY
elif line.startswith("##"):
line = line.lower()
if line.startswith("##fileformat"):
self.type = vcfKey.FIRSTLINE
elif line.startswith("##contig"):
self.type = vcfKey.CONTIG
contigID = line[line.find("ID=") + 3 :]
contigID = contigID[: contigID.find(",")]
self.chromosome = standardizeChromosome(contigID)
else:
self.type = vcfKey.OTHER_META
self.line = line
elif line.startswith("#"):
self.type = vcfKey.HEADER
else:
self.type = vcfKey.REGULAR
temp = vcfLine(line.strip().split("\t"))
temp.extractChrAndPos()
self.chromosome = temp.chromosome
self.position = temp.position
def run(args):
separateInfoFields = args.separate_info_fields.strip().lower() == "true"
countSeparate = args.count_separate.strip().lower() == "true"
ignoreFields = args.ignore_fields
if ignoreFields == None:
ignoreFields = []
ignoreFields = set(ignoreFields)
posLengthWarned = False
allChrs = []
positions = []
alleleColumn = infoDetails("Ref/Alt", 1, False)
qualColumn = infoDetails("QUAL", 1, False)
filterColumn = infoDetails("FILTER", args.max_strings, False)
infoFields = {"Ref/Alt":alleleColumn,"QUAL":qualColumn,"FILTER":filterColumn}
# TODO: get the numeric ranges, all valid categorical values
infile = open(args.infile,'r')
for line in infile:
line = line.strip()
if len(line) <= 1:
continue
elif line.startswith("#"):
temp = line.lower()
if temp.startswith("##info"):
newTag = line[temp.find("id=")+3:]
newTag = newTag[:newTag.find(',')]
if infoFields.has_key(newTag):
raise Exception("Duplicate INFO ID or use of reserved ID:\t%s" % newTag)
infoFields[newTag] = infoDetails(newTag, args.max_strings, countSeparate)
if newTag in ignoreFields:
infoFields[newTag].maxedOut = True
elif temp.startswith("##filter"):
newTag = line[temp.find("id=")+3:]
newTag = newTag[:newTag.find(',')]
filterColumn.addCategory(newTag, 0)
elif temp.startswith("##contig"):
chrom = line[temp.find("id=")+3:]
chrom = newTag[:chrom.find(',')]
chrom = standardizeChromosome(chrom)
chrLength = line[temp.find("length=")+3:]
chrLength = chrLength[:chrLength.find(',')]
allChrs.append(chrom)
positions.append((0,int(chrLength)))
else:
# a sneaky way of freezing the filter column; if other filters are added (without a .vcf pragma line) or we aren't separating info fields,
# other strings will make this column max out early
filterColumn.maxCategories = len(filterColumn.categories[0])
else:
line = vcfLine(line.split('\t'))
line.extractChrAndPos()
if not line.chromosome not in allChrs:
allChrs.append(line.chromosome)
positions.append((0,0))
chrIndex = allChrs.index(line.chromosome)
if line.position > positions[chrIndex][1]:
positions[chrIndex] = (0,line.position)
if not posLengthWarned:
sys.stderr.write('WARNING: Either ##contig pragma lines aren\'t supplied in your .vcf file or a variant has a position beyond the length of a chromosome.')
sys.stderr.write(' In either case, be aware that chromosome lengths in the .cvf file may not be accurate.')
posLengthWarned = True
line.extractAlleles()
alleles = line.alleles
if not separateInfoFields:
alleles = ",".join(alleles)
alleleColumn.addArbitraryValue(line.alleles)
line.extractQual()
qualColumn.addArbitraryValue(line.qual)
line.extractFilters()
filters = line.filters
if not separateInfoFields:
filters = ",".join(filters)
filterColumn.addArbitraryValue(filters)
line.extractInfo()
for k,v in line.info.iteritems():
if not infoFields.has_key(k):
raise Exception("Missing ##INFO pragma for: %s" % k)
if separateInfoFields:
v = ",".split(v)
infoFields[k].addArbitraryValue(v)
infile.close()
print "Creating file..."
outfile = open(args.outfile, 'w')
outfile.write("##\t%s created from %s on %s\n" % (args.outfile,args.infile,str(datetime.datetime.now())))
outfile.write("#\tChromosome\tCHR\t%s\n" % ("\t".join(allChrs)))
outfile.write("#\tPosition\tPOS\t%s\n" % ("\t".join(["(%i,%i)" % p for p in positions])))
outfile.write("#\tID\tID\n")
headers = []
fieldOrder = sorted(infoFields.iterkeys())
for f in fieldOrder:
pragmas = infoFields[f].getPragmas()
for p in pragmas:
outfile.write(p + "\n")
h = p.split("\t")[1]
headers.append(h)
outfile.write('Chromosome\tPosition\tID\t%s\n' % ("\t".join(headers)))
infile = open(args.infile,'r')
for line in infile:
line = line.strip()
if len(line) <= 1 or line.startswith("#"):
continue
line = vcfLine(line.split('\t'))
line.extractChrAndPos()
line.extractInfo()
line.extractAlleles()
line.info["Ref/Alt"] = line.alleles
line.extractQual()
line.info["QUAL"] = str(line.qual)
line.extractFilters()
line.info["FILTER"] = line.filters
outfile.write("%s\t%i\t%s" % (line.chromosome,line.position,line.name))
for f in fieldOrder:
values = line.info[f]
if isinstance(values,list):
if separateInfoFields:
values = "\t".join(values)
else:
values = ",".join(values)
outfile.write("\t%s" % values)
outfile.write("\n")
infile.close()
outfile.close()
def run(args):
infile = open(args.infile, "r")
outfile = open(args.outfile, "w")
failfile = None
if args.failfile != "":
failfile = open(args.failfile, "w")
errfile = None
if args.errfile != "":
errfile = open(args.errfile, "w")
if args.expression != "":
tempfile = open(args.expression, "r")
expression = tempfile.readline()
tempfile.close()
else:
expression = "True"
columns = []
if args.columns != "":
tempfile = open(args.columns, "r")
for line in tempfile:
line = line.strip()
columns.append(line)
tempfile.close()
bedRegions = None
if args.bed != "":
bedRegions = []
tempfile = open(args.bed, "r")
for line in tempfile:
bedRegions.append(bedLine(line.split()))
tempfile.close()
for line in infile:
if len(line) <= 1:
continue
elif line.startswith("#"):
outfile.write(line)
if failfile != None:
failfile.write(line)
if errfile != None:
errfile.write(line)
continue
else:
line = vcfLine(line.strip().split("\t"))
expArgs = []
for c in columns:
if c == "CHROM":
line.extractChrAndPos()
expArgs.append(line.chromosome)
elif c == "POS":
line.extractChrAndPos()
expArgs.append(line.position)
elif c == "ID":
line.extractChrAndPos()
expArgs.append(line.name)
elif c == "QUAL":
line.extractQual()
expArgs.append(line.qual)
elif c == "FILTER":
line.extractFilters()
expArgs.append(line.filters)
else:
line.extractInfo()
expArgs.append(line.info.get(c, "."))
# first see if it fails the .bed regions
if bedRegions != None:
passedBed = False
line.extractChrAndPos()
for bed in bedRegions:
if bed.contains(line.chromosome, line.position):
passedBed = True
break
if not passedBed:
if failfile != None:
failfile.write(str(line))
continue
exp = expression % tuple(expArgs)
try:
result = eval(exp)
if result == True:
outfile.write(str(line))
elif result == False:
if failfile != None:
failfile.write(str(line))
else:
if errfile != None:
errfile.write(str(line))
except:
if errfile != None:
errfile.write(str(line))
infile.close()
outfile.close()
if errfile != None:
errfile.close()
def run(args):
delimiter,headers,chromColumn,posColumn,idColumn = sniffCsv(args.csvfile)
if args.exact == None and args.nearest == None and args.interpolate == None:
temp = headers
temp.remove('CHROM')
temp.remove('POS')
args.exact = temp
exact = {}
nearest = {}
interpolate = {}
vcffile = open(args.infile,'r')
csvfile = open(args.csvfile,'r')
outfile = open(args.outfile,'w')
csvbasename = os.path.split(args.csvfile)[1]
takenTags = set()
vcfHeaderLine = ""
while True:
line = vcffile.readline()
if len(line) <= 1:
continue
elif line.startswith("##"):
if line.startswith("##INFO"):
newTag = line[line.find("ID=")+3:]
newTag = newTag[:newTag.find(',')]
takenTags.add(newTag)
outfile.write(line)
elif line.startswith("#"):
vcfHeaderLine = line
break
else:
raise Exception("Missing a header line or something else is wrong...")
if args.exact != None:
for x in args.exact:
if not x in headers:
raise Exception('Column header "%s" doesn\'t exist in %s' % (x,csvbasename))
temp = x
dupNumber = 2
while x in takenTags:
x = temp + str(dupNumber)
dupNumber += 1
takenTags.add(x)
outfile.write('##INFO=<ID=%s,Number=.,Type=String,Description=\"addCSVtoVCF.py: Column %s from %s in --exact match mode\">\n' % (x,temp,csvbasename))
exact[x] = headers.index(x)
if args.nearest != None:
for x in args.nearest:
if not x in headers:
raise Exception('Column header "%s" doesn\'t exist in %s' % (x,args.csvfile))
temp = x
dupNumber = 2
while x in takenTags:
x = temp + str(dupNumber)
dupNumber += 1
takenTags.add(x)
outfile.write('##INFO=<ID=%s,Number=.,Type=String,Description=\"addCSVtoVCF.py: Column %s from %s in --nearest match mode\">\n' % (x,temp,csvbasename))
nearest[x] = headers.index(x)
if args.interpolate != None:
for x in args.interpolate:
if not x in headers:
raise Exception('Column header "%s" doesn\'t exist in %s' % (x,args.csvfile))
temp = x
dupNumber = 2
while x in takenTags:
x = temp + str(dupNumber)
dupNumber += 1
takenTags.add(x)
outfile.write('##INFO=<ID=%s,Number=.,Type=String,Description=\"addCSVtoVCF.py: Column %s from %s in --interpolate match mode\">\n' % (x,temp,csvbasename))
interpolate[x] = headers.index(x)
outfile.write(vcfHeaderLine)
# grab our first lines
vLine = vcffile.readline()
vLine = vcfLine(vLine.strip().split('\t'))
vLine.extractChrAndPos()
vLine.extractInfo()
lastCline = None
csvfile.readline() # skip the header
cLine = csvLine(csvfile.readline().strip().split(delimiter),chromColumn,posColumn,idColumn)
while True:
speedAhead = cLine != None # a flag that lets us just spit out .vcf lines because we know that either the .csv file has finished or there are no new .csv lines on the same chromosome
# ... Are we even on the same chromosome?
while speedAhead and cLine.chrom != vLine.chromosome:
# the .csv file is ahead by a whole chromosome at least... keep lastCline intact and just spew out .vcf lines until it catches up
if chromosomeOrder.index(cLine.chrom) > chromosomeOrder.index(vLine.chromosome):
speedAhead = False
break
else:
# okay, the .csv file is behind the .vcf file by at least a chromsome... speed ahead until we catch up or run out of .csv data
lastCline = cLine
cLine = csvfile.readline()
if not cLine:
# shoot... we're out of .csv data. We already know that lastCline wasn't on the same chromosome as the current vLine, so make it None as well
cLine = None
lastCline = None
speedAhead = False
break
else:
cLine = csvLine(cLine.strip().split(delimiter),chromColumn,posColumn,idColumn)
# Okay, now we're on the same chromosome... zip ahead until cLine and lastCline are straddling vLine
while speedAhead and cLine.pos < vLine.position:
lastCline = cLine
cLine = csvfile.readline()
if not cLine:
# shoot... out of .csv data. We know lastCline is still on the same chromosome, so preserve that, but make cLine None so we know nothing is left
cLine = None
break
else:
cLine = csvLine(cLine.strip().split(delimiter),chromColumn,posColumn,idColumn)
# Whew! We're finally straddling the vLine...
# Check the super-special case first (exact match)
if cLine != None and cLine.pos == vLine.position:
for x,i in exact.iteritems():
vLine.info[x] = cLine.columns[i]
for x,i in nearest.iteritems():
vLine.info[x] = cLine.columns[i]
for x,i in interpolate.iteritems():
vLine.info[x] = cLine.columns[i]
elif cLine != None: # cLine.pos will be > vLine.position
if not args.omit_mismatches:
for x,i in exact.iteritems():
vLine.info[x] = "."
if lastCline == None:
for x,i in nearest.iteritems():
vLine.info[x] = cLine.columns[i]
for x,i in interpolate.iteritems():
vLine.info[x] = cLine.columns[i]
else:
closestLine = lastCline if vLine.position - lastCline.pos <= cLine.pos - vLine.position else cLine
for x,i in nearest.iteritems():
vLine.info[x] = closestLine.columns[i]
for x,i in interpolate.iteritems():
try:
lastVal = float(lastCline.columns[i])
nextVal = float(cLine.columns[i])
vLine.info[x] = str(lastVal + (nextVal - lastVal)*(vLine.position - lastCline.pos)/(cLine.pos - lastCline.pos))
except ValueError:
vLine.info[x] = closestLine.columns[i]
else: # cLine == None
if lastCline == None:
if not args.omit_mismatches:
for x,i in exact.iteritems():
vLine.info[x] = "."
for x,i in nearest.iteritems():
vLine.info[x] = "."
for x,i in interpolate.iteritems():
vLine.info[x] = "."
else:
if not args.omit_mismatches:
for x,i in exact.iteritems():
vLine.info[x] = "."
for x,i in nearest.iteritems():
vLine.info[x] = lastCline.columns[i]
for x,i in interpolate.iteritems():
vLine.info[x] = lastCline.columns[i]
# Okay, we've copied over everything; write the line
outfile.write(str(vLine))
# Grab the next one
vLine = vcffile.readline()
if not vLine:
break # No more variants - we're done!
vLine = vcfLine(vLine.strip().split('\t'))
vLine.extractChrAndPos()
vLine.extractInfo()
csvfile.close()
vcffile.close()
outfile.close()
for line in infile:
if len(line) <= 1:
# Skip blank lines
continue
elif line.startswith('#'):
# Skip header pragma lines
if line.lower().startswith('#chrom'):
# Get all the individual IDs from the header, write it to the file
individualIDs = line.strip().split('\t')[9:]
outfile.write(','.join(individualIDs))
outfile.write('\n')
else:
assert individualIDs != None # In a well-formed .vcf file, we'll have run across the header line before any data
# Here is where my library comes in - we first build a vcfLine object with the columns in the .vcf file
line = vcfLine(line.strip().split('\t'))
# Ideally, these two steps would be performed automatically, but sometimes we might want
# to skip them to save time in practice:
# We have to first extract the alleles before I can reference line.alleles
line.extractAlleles()
# We have to first extract all genotypes before I can reference line.genotypes
line.extractGenotypes()
for i in xrange(len(individualIDs)):
allele1,allele2,phased,attributes = line.genotypes[i]
outfile.write(line.alleles[allele1]) # write the first allele letters - to write the number, just write allele1
outfile.write("|" if phased else "/")
outfile.write(line.alleles[allele2]) # write the second allele letters - to write the number, just write allele2
for line in infile:
if len(line) <= 1:
# Skip blank lines
continue
elif line.startswith('#'):
# Skip header pragma lines
if line.lower().startswith('#chrom'):
# Get all the individual IDs from the header, write it to the file
individualIDs = line.strip().split('\t')[9:]
outfile.write(','.join(individualIDs))
outfile.write('\n')
else:
assert individualIDs != None # In a well-formed .vcf file, we'll have run across the header line before any data
# Here is where my library comes in - we first build a vcfLine object with the columns in the .vcf file
line = vcfLine(line.strip().split('\t'))
# Ideally, these two steps would be performed automatically, but sometimes we might want
# to skip them to save time in practice:
# We have to first extract the alleles before I can reference line.alleles
line.extractAlleles()
# We have to first extract all genotypes before I can reference line.genotypes
line.extractGenotypes()
for i in xrange(len(individualIDs)):
allele1, allele2, phased, attributes = line.genotypes[i]
outfile.write(
line.alleles[allele1]
) # write the first allele letters - to write the number, just write allele1
def run(args):
separateInfoFields = args.separate_info_fields.strip().lower() == "true"
numberAlleles = args.numbered_alleles.strip().lower() == "true"
includeGenotypes = args.include_genotypes.strip().lower() == "true"
includeGenotypeAttributes = args.include_genotype_attributes.strip().lower() == "true"
if includeGenotypeAttributes:
includeGenotypes = True
ignoreFields = args.ignore_fields
if ignoreFields == None:
ignoreFields = set([])
ignoreFields = set(ignoreFields)
numAltAlleles = 1
numFilters = 1
infoOrder = []
infoHeaders = {}
peopleOrder = []
formatOrder = []
formatHeaders = {}
infile = open(args.infile,'r')
for line in infile:
line = line.strip()
if len(line) <= 1:
continue
elif line.startswith("#"):
temp = line.lower()
if temp.startswith("##info"):
newTag = line[temp.find("id=")+3:]
newTag = newTag[:newTag.find(',')]
if not newTag in ignoreFields:
if not infoHeaders.has_key(newTag):
infoOrder.append(newTag)
infoHeaders[newTag] = 1
elif temp.startswith("##format"):
newTag = line[temp.find("id=")+3:]
newTag = newTag[:newTag.find(',')]
if not newTag in ignoreFields and not newTag == 'GT':
if not formatHeaders.has_key(newTag):
formatOrder.append(newTag)
formatHeaders[newTag] = 1
elif temp.startswith("#chrom"):
peopleOrder = line.split('\t')[9:]
else:
line = vcfLine(line.split('\t'))
if separateInfoFields:
line.extractAlleles()
numAltAlleles = max(numAltAlleles,len(line.alleles)-1) # don't include the REF allele
line.extractFilters()
numFilters = max(numFilters,len(line.filters))
line.extractInfo()
for k,v in line.info.iteritems():
if not infoHeaders.has_key(k):
infoOrder.append(newTag)
infoHeaders[k] = 1
if separateInfoFields and isinstance(v,list):
infoHeaders[k] = max(infoHeaders[k],len(v))
if includeGenotypeAttributes:
line.extractFormat()
line.extractGenotypes()
for i,p in enumerate(peopleOrder):
allele0,allele1,phased,attrs = line.genotypes[i] # @UnusedVariable
for j,f in enumerate(line.format[1:]):
if len(attrs) > j:
formatHeaders[f] = max(formatHeaders[f],len(attrs[j].split(',')))
infile.close()
print "Creating file..."
outfile = open(args.outfile, 'w')
outfile.write('Chromosome\tPosition\tID\tReference_Allele')
if separateInfoFields and numAltAlleles > 1:
for x in xrange(numAltAlleles):
outfile.write('\tAlternate_Allele_%i' % (x+1))
else:
outfile.write('\tAlternate_Allele')
outfile.write('\tQual')
if separateInfoFields and numFilters > 1:
for x in xrange(numFilters):
outfile.write('\tFilter_%i' % (x+1))
else:
outfile.write('\tFilter')
for i in infoOrder:
if separateInfoFields and infoHeaders[i] > 1:
for x in xrange(infoHeaders[i]):
outfile.write('\t%s_%i' % (i,x+1))
else:
outfile.write('\t%s' % i)
if includeGenotypes:
for p in peopleOrder:
outfile.write('\t%s_Allele_1' % p)
outfile.write('\t%s_Allele_2' % p)
if includeGenotypeAttributes:
outfile.write('\t%s_Phased' % p)
for f in formatOrder:
if separateInfoFields and formatHeaders[f] > 1:
for x in xrange(formatHeaders[f]):
outfile.write('\t%s_%s_%i' % (p,f,x+1))
else:
outfile.write('\t%s_%s' % (p,f))
outfile.write('\n')
infile = open(args.infile,'r')
for line in infile:
line = line.strip()
if len(line) <= 1 or line.startswith("#"):
continue
line = vcfLine(line.split('\t'))
line.extractChrAndPos()
outfile.write("%s\t%i\t%s" % (line.chromosome,line.position,line.name))
line.extractAlleles()
outfile.write('\t%s' % line.alleles[0])
if separateInfoFields:
for a in line.alleles[1:]:
outfile.write('\t%s' % a)
x = len(line.alleles)-1
while x < numAltAlleles:
outfile.write('\t')
x += 1
else:
outfile.write('\t%s' % ','.join(line.alleles[1:]))
line.extractQual()
outfile.write("\t%f" % line.qual)
line.extractFilters()
if separateInfoFields:
for f in line.filters:
outfile.write('\t%s' % f)
x = len(line.filters)
while x < numFilters:
outfile.write('\t')
x += 1
else:
outfile.write('\t%s' % ','.join(line.filters))
line.extractInfo()
for i in infoOrder:
if not line.info.has_key(i):
if separateInfoFields:
for x in xrange(infoHeaders[i]):
outfile.write('\t')
else:
outfile.write('\t')
else:
values = line.info[i]
if not isinstance(values,list):
values = [values]
for j,v in enumerate(values):
if v == None:
values[j] = i
if separateInfoFields:
for v in values:
outfile.write('\t%s' % v)
x = len(values)
while x < infoHeaders[i]:
outfile.write('\t')
x += 1
else:
outfile.write('\t%s' % ','.join(values))
if includeGenotypes:
line.extractFormat()
line.extractGenotypes()
for i,p in enumerate(peopleOrder):
allele0,allele1,phased,attrs = line.genotypes[i]
if allele0 == None:
allele0 = '.'
elif not numberAlleles:
allele0 = line.alleles[allele0]
if allele1 == None:
allele1 = '.'
elif not numberAlleles:
allele1 = line.alleles[allele1]
outfile.write('\t%s\t%s' % (allele0,allele1))
if includeGenotypeAttributes:
outfile.write('\t%s' % ('Y' if phased else 'N'))
for j,f in enumerate(formatOrder):
if not f in line.format:
if separateInfoFields:
for x in xrange(formatHeaders[f]):
outfile.write('\t')
else:
outfile.write('\t')
else:
attrIndex = line.format.index(f)-1
if attrIndex >= len(attrs):
if separateInfoFields:
for x in xrange(formatHeaders[f]):
outfile.write('\t')
else:
outfile.write('\t')
else:
values = attrs[attrIndex].split(',')
if separateInfoFields:
for v in values:
outfile.write('\t%s' % v)
x = len(values)
while x < formatHeaders[f]:
outfile.write('\t')
x += 1
else:
outfile.write('\t%s' % ','.join(values))
outfile.write("\n")
infile.close()
outfile.close()
def run(args):
separateInfoFields = args.separate_info_fields.strip().lower() == "true"
countSeparate = args.count_separate.strip().lower() == "true"
ignoreFields = args.ignore_fields
if ignoreFields == None:
ignoreFields = []
ignoreFields = set(ignoreFields)
posLengthWarned = False
allChrs = []
positions = []
alleleColumn = infoDetails("Ref/Alt", 1, False)
qualColumn = infoDetails("QUAL", 1, False)
filterColumn = infoDetails("FILTER", args.max_strings, False)
infoFields = {
"Ref/Alt": alleleColumn,
"QUAL": qualColumn,
"FILTER": filterColumn
}
# TODO: get the numeric ranges, all valid categorical values
infile = open(args.infile, 'r')
for line in infile:
line = line.strip()
if len(line) <= 1:
continue
elif line.startswith("#"):
temp = line.lower()
if temp.startswith("##info"):
newTag = line[temp.find("id=") + 3:]
newTag = newTag[:newTag.find(',')]
if infoFields.has_key(newTag):
raise Exception(
"Duplicate INFO ID or use of reserved ID:\t%s" %
newTag)
infoFields[newTag] = infoDetails(newTag, args.max_strings,
countSeparate)
if newTag in ignoreFields:
infoFields[newTag].maxedOut = True
elif temp.startswith("##filter"):
newTag = line[temp.find("id=") + 3:]
newTag = newTag[:newTag.find(',')]
filterColumn.addCategory(newTag, 0)
elif temp.startswith("##contig"):
chrom = line[temp.find("id=") + 3:]
chrom = newTag[:chrom.find(',')]
chrom = standardizeChromosome(chrom)
chrLength = line[temp.find("length=") + 3:]
chrLength = chrLength[:chrLength.find(',')]
allChrs.append(chrom)
positions.append((0, int(chrLength)))
else:
# a sneaky way of freezing the filter column; if other filters are added (without a .vcf pragma line) or we aren't separating info fields,
# other strings will make this column max out early
filterColumn.maxCategories = len(filterColumn.categories[0])
else:
line = vcfLine(line.split('\t'))
line.extractChrAndPos()
if not line.chromosome not in allChrs:
allChrs.append(line.chromosome)
positions.append((0, 0))
chrIndex = allChrs.index(line.chromosome)
if line.position > positions[chrIndex][1]:
positions[chrIndex] = (0, line.position)
if not posLengthWarned:
sys.stderr.write(
'WARNING: Either ##contig pragma lines aren\'t supplied in your .vcf file or a variant has a position beyond the length of a chromosome.'
)
sys.stderr.write(
' In either case, be aware that chromosome lengths in the .cvf file may not be accurate.'
)
posLengthWarned = True
line.extractAlleles()
alleles = line.alleles
if not separateInfoFields:
alleles = ",".join(alleles)
alleleColumn.addArbitraryValue(line.alleles)
line.extractQual()
qualColumn.addArbitraryValue(line.qual)
line.extractFilters()
filters = line.filters
if not separateInfoFields:
filters = ",".join(filters)
filterColumn.addArbitraryValue(filters)
line.extractInfo()
for k, v in line.info.iteritems():
if not infoFields.has_key(k):
raise Exception("Missing ##INFO pragma for: %s" % k)
if separateInfoFields:
v = ",".split(v)
infoFields[k].addArbitraryValue(v)
infile.close()
print "Creating file..."
outfile = open(args.outfile, 'w')
outfile.write("##\t%s created from %s on %s\n" %
(args.outfile, args.infile, str(datetime.datetime.now())))
outfile.write("#\tChromosome\tCHR\t%s\n" % ("\t".join(allChrs)))
outfile.write("#\tPosition\tPOS\t%s\n" %
("\t".join(["(%i,%i)" % p for p in positions])))
outfile.write("#\tID\tID\n")
headers = []
fieldOrder = sorted(infoFields.iterkeys())
for f in fieldOrder:
pragmas = infoFields[f].getPragmas()
for p in pragmas:
outfile.write(p + "\n")
h = p.split("\t")[1]
headers.append(h)
outfile.write('Chromosome\tPosition\tID\t%s\n' % ("\t".join(headers)))
infile = open(args.infile, 'r')
for line in infile:
line = line.strip()
if len(line) <= 1 or line.startswith("#"):
continue
line = vcfLine(line.split('\t'))
line.extractChrAndPos()
line.extractInfo()
line.extractAlleles()
line.info["Ref/Alt"] = line.alleles
line.extractQual()
line.info["QUAL"] = str(line.qual)
line.extractFilters()
line.info["FILTER"] = line.filters
outfile.write("%s\t%i\t%s" %
(line.chromosome, line.position, line.name))
for f in fieldOrder:
values = line.info[f]
if isinstance(values, list):
if separateInfoFields:
values = "\t".join(values)
else:
values = ",".join(values)
outfile.write("\t%s" % values)
outfile.write("\n")
infile.close()
outfile.close()