def dprimecalc(record1, record2, snpcheck = True):
'''Calculates Lewontin's D' statistic (D/Dmax) between two VCF records.
Will check that records are biallelic (single-ALT) SNPs unless snpcheck = False.
'''
if snpcheck:
snpchecker(record1, record2)
elif not snpcheck:
pass
values = freqsgetter(record1, record2, snpcheck = False)[1] # get allele frequencies
d = dcalc(record1, record2, snpcheck = False)
if d >= 0:
dmax = min(values['p1'] * values['q2'], values['p2'] * values['q1'])
if dmax == 0:
out = 0
else:
out = d/dmax
elif d < 0:
dmin = max(-1 * values['p1'] * values['q1'], -1 * values['p2'] * values['q2'])
if dmin == 0:
out = 0
else:
out = d/dmin
elif round(d, 6) == 0:
out = 0
return out
def quick_dcalc(record1, record2):
haps = freqsgetter(record1, record2, snpcheck = False)[2]
try:
LHS = haps['AB'] * haps['ab']
except KeyError: # either hap missing
LHS = 0
try:
RHS = haps['Ab'] * haps['aB']
except KeyError:
RHS = 0
d = LHS - RHS
# d = round(d, 5)
return d
def freqscalc(record1, record2, snpcheck = True, aaf = False):
'''Exploratory convenience function. Given two VCF records, returns
observed haplotype frequencies. Will check that records are biallelic (single-ALT) SNPs
unless snpcheck = False. aaf will return AF values hardcoded in the VCF
itself, while aaf = False (default) will make freqscalc calculate them instead
(more accurate option).
'''
if snpcheck:
snpchecker(record1, record2)
elif not snpcheck:
pass
# get allele frequencies
if aaf == True:
p = 1 - record1.aaf[0]
q = 1 - record2.aaf[0]
p2 = record1.aaf[0]
q2 = record2.aaf[0]
elif aaf == False:
values = freqsgetter(record1, record2)[1]
p = values['p1']
q = values['q1']
p2 = values['p2']
q2 = values['q2']
print(record1.CHROM, record1.POS, '- ref', record1.REF, 'alt', record1.ALT[0])
print(record2.CHROM, record2.POS, '- ref', record2.REF, 'alt', record2.ALT[0])
print('p1 ', p, 'p2 ', p2)
print('q1 ', q, 'q2 ', q2)
# get samples + check for same samples b/w both records
strainlist = straingetter(record1, record2)
# score haplotypes
haplist = []
for strain in strainlist:
gt1 = record1.genotype(strain)['GT']
gt2 = record2.genotype(strain)['GT']
if gt1 == '.' or gt2 == '.':
continue
if gt1 == '1' and gt2 == '1':
outgt = str(record1.ALT[0]) + str(record2.ALT[0])
elif gt1 == '1' and gt2 == '0':
outgt = str(record1.ALT[0]) + record2.REF
elif gt1 == '0' and gt2 == '1':
outgt = record1.REF + str(record2.ALT[0])
elif gt1 == '0' and gt2 == '0':
outgt = record1.REF + record2.REF
haplist.append(outgt) # create list of observed genotypes
uniques = set(haplist)
for hap in uniques:
print(hap, round(haplist.count(hap)/len(haplist), 5))
def r2calc(record1, record2, snpcheck = True):
'''Calculates r^2 (correlation) between two VCF records.
Will check that records are biallelic (single-ALT) SNPs unless snpcheck = False.
'''
if snpcheck:
snpchecker(record1, record2)
elif not snpcheck:
pass
values = freqsgetter(record1, record2, snpcheck = False)[1]
if values['p1'] == 0 or values['q1'] == 0 or values['p2'] == 0 or values['q2'] == 0:
out = 0
else:
dsquared = dcalc(record1, record2, snpcheck = False)**2
out = dsquared/(values['p1'] * values['q1'] * values['p2'] * values['q2'])
# out = round(out, 4)
return out
def dcalc(record1, record2, snpcheck = True):
'''Calculates D statistic between two VCF records.
Will check that records are biallelic (single-ALT) SNPs unless snpcheck = False.
'''
if snpcheck:
snpchecker(record1, record2)
elif not snpcheck:
pass
haps = freqsgetter(record1, record2, snpcheck = False)[2]
try:
LHS = haps['AB'] * haps['ab']
except KeyError: # either hap missing
LHS = 0
try:
RHS = haps['Ab'] * haps['aB']
except KeyError:
RHS = 0
d = LHS - RHS
# d = round(d, 5)
return d