def main():
usage = "usage: %prog [options] arg"
parser = OptionParser(usage)
parser.add_option("--info", type="string", dest="infotag", help="INFO tag id that annotates what type of variant the VCF record is", default="TYPE")
parser.add_option("--filter", type="string", dest="filter", help="only analyze records with matching filter (default is None)", default=None)
(options, args)=parser.parse_args()
if options.infotag == "":
sys.stderr.write("provide a value for --info parameter!\n")
exit(1)
variant_dict={} #key variant type value VcfRecord object
vcfilename=args[0]
vcfh=open(vcfilename,'r')
#instantiate a VcfFile object
vcfobj=VcfFile(vcfilename)
#parse its metainfo lines (ones that begin with ##)
vcfobj.parseMetaAndHeaderLines(vcfh)
descriptors = vcfobj.getMetaInfoDescription()
infoids=[]
for (tag, description) in descriptors:
tag
infoids.append(tag)
if options.infotag not in infoids and options.infotag != 'QUAL':
sys.stderr.write(options.infotag + " tag not in ##INFO headers!\n")
exit(1)
pattern=options.infotag+'=(\S+)'
for vrec in vcfobj.yieldVcfRecord(vcfh):
if vrec.getFilter() != options.filter and options.filter != None: continue
searchresult=re.search(pattern, vrec.getInfo() )
if re.search(pattern, vrec.getInfo() ) == None:
continue
else:
value=re.search(pattern, vrec.getInfo() ).groups()[0]
#rint value
if value not in variant_dict.keys():
variant_dict[value]=[]
variant_dict[value].append( vrec )
else:
variant_dict[value].append( vrec )
sum=0
sys.stderr.write("types and count of different variant classes in " + vcfilename + "\n")
for k in variant_dict.keys():
print k, len( variant_dict[k] )
sum+=len( variant_dict[k] )
print "TOTAL:", sum
def binned_bitsets_from_vcffile( vcfilename, chrom_col=0, start_col=1, upstream_pad=0, downstream_pad=0, lens={} ):
"""
Read a vcffile into a dictionary of bitsets. The defaults arguments
- 'vcfilename' should be a filename for vcf file
- 'chrom_col', 'start_col', and 'end_col' must exist in each line.
- if 'lens' is provided bitset sizes will be looked up from it, otherwise
chromosomes will be assumed to be the maximum size
- the bitset interval made into a zero-based, half-open interval!!!!!!!
"""
last_chrom = None
last_bitset = None
bitsets = dict()
MAX=2147483647
vcfobj=VcfFile(vcfilename)
fh=open(vcfilename,'r')
for vrec in vcfobj.yieldVcfRecord(fh):
filtercode = vrec.getFilter()
chrom = vrec.getChrom()
pos=int( vrec.getPos() )
#if filtercode != filtercodeoption and filtercodeoption != None:
# continue
if filtercode != 'PASS':
if filtercode == '.':
pass
else:
continue
chrom="chr"+chrom
if chrom != last_chrom:
if chrom not in bitsets:
if chrom in lens:
size = lens[chrom]
else:
size = MAX
bitsets[chrom] = BinnedBitSet( size )
last_chrom = chrom
last_bitset = bitsets[chrom]
start, end = (pos-1, pos)
if upstream_pad: start = max( 0, start - upstream_pad )
if downstream_pad: end = min( size, end + downstream_pad )
if start > end: warn( "Interval start after end!" )
last_bitset.set_range( start, end-start )
fh.close()
return bitsets
def main():
usage = "usage: %prog [options] arg"
parser = OptionParser(usage)
usage = "usage: %prog [options] file.vcf \n print summary information about site depth in records of a VCF file\n"
parser = OptionParser(usage)
parser.add_option("--max", type="int", dest="max", help="skip records that are greater than or equal to max (default sys.maxint)", default=sys.maxint)
#parser.add_option("--v", action="store_true", dest="snp", help="restrict analysis to SNPs (must have INFO ID SNP in header")
(options, args)=parser.parse_args()
vcfilename=args[0]
fileName, fileExtension = os.path.splitext(vcfilename)
#nuller.12:80717441..80717681.vcf
regionpattern='nuller.(\d+):(\d+)..(\d+)'
results=re.search(regionpattern,fileName ).groups()
regionstr="\t".join(list(results))
vcfh=open(vcfilename,'r')
#instantiate a VcfFile object
vcfobj=VcfFile(vcfilename)
#parse its metainfo lines (ones that begin with ##)
vcfobj.parseMetaLines(vcfh)
descriptors = vcfobj.getMetaInfoDescription()
infoids=[]
for (tag, description) in descriptors:
infoids.append(tag)
if 'DP' not in infoids:
sys.stderr.write("DP tag not in ##INFO headers!")
exit(1)
vcfh.seek(0)
vcfobj.parseHeaderLine(vcfh)
pattern='DP=(\d+)'
depth_list=[]
for vrec in vcfobj.yieldVcfRecord(vcfh):
dp=re.search(pattern, vrec.getInfo() ).groups()[0]
if dp == None:
sys.stderr.write("unable to parse DP value from INFO field\n")
continue
else:
if int(dp) >= options.max: continue
depth_list.append(int(dp))
maxDP=max( array (depth_list))
minDP= min (array (depth_list))
medianDP=median (array (depth_list))
meanDP=mean( array(depth_list))
length=len(depth_list)
outstr="\t".join([regionstr, str(maxDP), str(minDP), str(medianDP), str(meanDP), str(length)])
print outstr
