def do_work(fd_vcf, csv):
vcf = Vcf(fd_vcf)
vcf.load_meta_header()
grps = Group(csv)
#sys.stderr.write("# Of groups loaded: %s\n" % grps.num())
min_threshold = float(sys.argv[2])
min_num_all = float(sys.argv[3])
ComputeSnps(vcf, grps, min_threshold, min_num_all).run()
def doHeader(fd):
vcf = Vcf(fd)
vcf.load_meta_header()
ot_already_there = str(vcf.check_info('OT'))
if ot_already_there == True:
error("This vcf seems to have an OT INFO already. Bailing out.")
vcf.add_info('OT', '0', 'Flag', 'The site is on target.')
# print "Checking if INFO id=OT is there: " + str(vcf.check_info('OT'))
print vcf.get_meta()
return vcf
def doHeader(fd):
vcf = Vcf(fd)
vcf.load_meta_header()
ot_already_there = str(vcf.check_info('OT'))
if ot_already_there == True:
error("This vcf seems to have an OT INFO already. Bailing out.")
vcf.add_info('OT', '0', 'Flag', 'The site is on target.')
# print "Checking if INFO id=OT is there: " + str(vcf.check_info('OT'))
print vcf.get_meta()
return vcf
def doHeader(fd):
vcf = Vcf(fd)
vcf.load_meta_header()
ot_already_there = str(vcf.check_info('RDP'))
if ot_already_there == True:
error("This vcf seems to have an RDP INFO already. Bailing out.")
vcf.add_info('RDP', '1', 'Integer', 'Raw read coverage at locus.')
# print "Checking if INFO id=OT is there: " + str(vcf.check_info('OT'))
print vcf.get_meta()
return vcf
def do_work(fd_vcf, tsv_pheno, tsv_haplo): vcf = Vcf(fd_vcf) vcf.load_meta_header() grps_pheno, grps_haplo = Group(tsv_pheno), Group(tsv_haplo) matrix, a_sites, a_groups = prepare(vcf, grps_pheno, grps_haplo) print matrix.shape print a_groups #values = np.random.randn(100,100) * 10 #a_sites = ['1_100', '2_200', '3_300'] #a_groups = ['grp1', 'grp1', 'grp2'] cb_labels = ['HETE', 'HOMO_VAR', 'OTHER', 'NO_COVERAGE', 'HOMO_REF'] drdplots.Heatmap(matrix, cb_labels, a_sites, a_groups).plot()
def do_work(fd_vcf, tsv_pheno, tsv_haplo):
vcf = Vcf(fd_vcf)
vcf.load_meta_header()
grps_pheno, grps_haplo = Group(tsv_pheno), Group(tsv_haplo)
matrix, a_sites, a_groups = prepare(vcf, grps_pheno, grps_haplo)
print matrix.shape
print a_groups
#values = np.random.randn(100,100) * 10
#a_sites = ['1_100', '2_200', '3_300']
#a_groups = ['grp1', 'grp1', 'grp2']
cb_labels = ['HETE', 'HOMO_VAR', 'OTHER', 'NO_COVERAGE', 'HOMO_REF']
drdplots.Heatmap(matrix, cb_labels, a_sites, a_groups).plot()
def do_work(fd_vcf, min_num_samples):
vcf = Vcf(fd_vcf)
vcf.load_meta_header()
report(process_snps(vcf, min_num_samples))
class SnpFreq(object):
GENOME_SIZE = {'wgs': 3000000000, 'wes': 34000000}
MIN_QUAL = 20
def __init__(self, fd_vcf, exp_type, options):
self.fd_vcf = fd_vcf
self.exp_type = exp_type
self.drop = options.drop
self.list_s_snps = options.list_s_snps
self.__validate_type()
self.coordinates_in_file = options.coordinates_in_file
self.__load_vcf()
if options.coor_fn:
self.coor_fn = options.coor_fn
self.__load_species_snp_coordinates()
def __load_species_snp_coordinates(self):
fd = drdcommon.xopen(self.coor_fn)
d = {}
self.d_species_coor = d
n = 0
for l in fd:
n += 1
chrm, coor = l.split()
if not d.has_key(chrm):
d[chrm] = {}
d[chrm][int(coor)] = 1
fd.close()
logging.info("# of coordinates loaded: %d" % n)
logging.info("current memory usage in %dkb" % drdcommon.memory_usage())
def __load_vcf(self):
self.vcf = Vcf(self.fd_vcf)
self.vcf.load_meta_header()
if self.drop and (not self.coordinates_in_file
and self.vcf.num_of_samples < 2):
drdcommon.error(
"I need a population level vcf in order to drop species snps.")
def __validate_type(self):
if self.exp_type != 'wgs' and \
self.exp_type != 'wes' and \
self.exp_type != 'null':
raise_it('Invalid experiment type. Valid types: wgs or wes')
def __list_species_snps(self):
for l in self.vcf.each_snp():
snp = VcfSnp(l)
if snp.is_a_substitution() and \
snp.has_high_quality(self.MIN_QUAL) and \
snp.species_snp():
print(snp.coordinate(' '))
def __is_a_species_snp(self, snp):
if self.coordinates_in_file:
ch, co = snp.coordinate(' ').split()
return self.d_species_coor.has_key(
ch) and self.d_species_coor[ch].has_key(int(co))
else:
return snp.species_snp()
def __calculate_snp_freq(self):
"""
Compute the snp frequency (# of snps per kbp)
Drop snps that are indels, have low quality
If wes, also drop non coding regions
If drop is True, we have to drop species snps
"""
num_snps = 0
total = 0
for l in self.vcf.each_snp():
snp = VcfSnp(l)
total += 1
if snp.is_a_substitution() and snp.has_high_quality(self.MIN_QUAL):
if self.exp_type == 'wgs':
if not self.drop or (self.drop
and not self.__is_a_species_snp(snp)):
num_snps += 1
if self.exp_type == 'wes' and snp.in_coding_region():
if not self.drop or (self.drop
and not self.__is_a_species_snp(snp)):
num_snps += 1
logging.info("Total/counted: %d/%d" % (total, num_snps))
return (float(num_snps) / self.GENOME_SIZE[self.exp_type]) * 1000
def run(self):
if self.list_s_snps:
self.__list_species_snps()
else:
return self.__calculate_snp_freq()
class SnpFreq(object):
GENOME_SIZE = {'wgs': 3000000000, 'wes': 34000000}
MIN_QUAL = 20
def __init__(self, fd_vcf, exp_type, options):
self.fd_vcf = fd_vcf
self.exp_type = exp_type
self.drop = options.drop
self.list_s_snps = options.list_s_snps
self.__validate_type()
self.coordinates_in_file = options.coordinates_in_file
self.__load_vcf()
if options.coor_fn:
self.coor_fn = options.coor_fn
self.__load_species_snp_coordinates()
def __load_species_snp_coordinates(self):
fd = drdcommon.xopen(self.coor_fn)
d = {}
self.d_species_coor = d
n = 0
for l in fd:
n += 1
chrm, coor = l.split()
if not d.has_key(chrm):
d[chrm] = {}
d[chrm][int(coor)] = 1
fd.close()
logging.info("# of coordinates loaded: %d" % n)
logging.info("current memory usage in %dkb" % drdcommon.memory_usage())
def __load_vcf(self):
self.vcf = Vcf(self.fd_vcf)
self.vcf.load_meta_header()
if self.drop and (not self.coordinates_in_file and self.vcf.num_of_samples < 2):
drdcommon.error("I need a population level vcf in order to drop species snps.")
def __validate_type(self):
if self.exp_type != 'wgs' and \
self.exp_type != 'wes' and \
self.exp_type != 'null':
raise_it('Invalid experiment type. Valid types: wgs or wes')
def __list_species_snps(self):
for l in self.vcf.each_snp():
snp = VcfSnp(l)
if snp.is_a_substitution() and \
snp.has_high_quality(self.MIN_QUAL) and \
snp.species_snp():
print(snp.coordinate(' '))
def __is_a_species_snp(self, snp):
if self.coordinates_in_file:
ch, co = snp.coordinate(' ').split()
return self.d_species_coor.has_key(ch) and self.d_species_coor[ch].has_key(int(co))
else:
return snp.species_snp()
def __calculate_snp_freq(self):
"""
Compute the snp frequency (# of snps per kbp)
Drop snps that are indels, have low quality
If wes, also drop non coding regions
If drop is True, we have to drop species snps
"""
num_snps = 0
total = 0
for l in self.vcf.each_snp():
snp = VcfSnp(l)
total += 1
if snp.is_a_substitution() and snp.has_high_quality(self.MIN_QUAL):
if self.exp_type == 'wgs':
if not self.drop or (self.drop and not self.__is_a_species_snp(snp)):
num_snps += 1
if self.exp_type == 'wes' and snp.in_coding_region():
if not self.drop or (self.drop and not self.__is_a_species_snp(snp)):
num_snps += 1
logging.info("Total/counted: %d/%d" % (total, num_snps))
return (float(num_snps)/self.GENOME_SIZE[self.exp_type])*1000
def run(self):
if self.list_s_snps:
self.__list_species_snps()
else:
return self.__calculate_snp_freq()