heterocommon=0
p = [0.2, 0.2, 0.2]
notes= [defaultdict(lambda:0), defaultdict(lambda:0), defaultdict(lambda:0)]
errprob=0.1
for line in file('snplist'):
[ch, pos, snp, val, prob, qual, f] = line.split('\t')
prob=float(prob)
qual=float(qual)
if prob==0:
continue
if snp not in yaniv:
no23me += 1
continue
if len(yaniv[snp])==1 or yaniv[snp][0]==yaniv[snp][1]:
to_update=[0,2]
else:
to_update=[1,2]
for i in to_update:
p[i] = update(val, yaniv[snp], p[i], prob, notes[i], i==0, error_rate=.05)
print 'homozyg hits=%d, heterozyg hits=%d, sex chromosome hits=%d, h**o misses=%d hetero misses=%d' % (notes[2]['homohit'], notes[2]['heterohit'], notes[2]['sexhits'], notes[2]['homomiss'], notes[2]['heteromiss'])
heterocommon=notes[2]['hetcom']
heterorare=notes[2]['hetrare']
print 'hetero common=%d, hetero rare=%d, p=%.4f' % (heterocommon, heterorare, binom.cdf(min(heterocommon, heterorare), heterocommon+heterorare, 0.5))
print 'Skipping hetero: p(Yaniv)=%.2f%%' % (100*(p[0]))
print 'Using only hetero: p(Yaniv)=%.2f%%' % (100*(p[1]))
print 'Using everything: p(Yaniv)=%.2f%%' % (100*(p[2]))
for line in file('/windows/ALL.chr6.ours.vcf'):
words = line.rstrip().split('\t')
snp = words[2]
if snp not in ourvals:
continue
meaning = words[3:5]
if len(meaning[1]) > 1:
meaning = [meaning[0]] + meaning[1].split(',')
for i in range(ngenomes):
ref = words[i + 9]
try:
ref = meaning[int(ref[0])] + meaning[int(ref[2])]
except IndexError as e:
print ref
print meaning
throw(e)
ps[i] = update(ourvals[snp], ref, ps[i], ourps[snp], notes[i])
results = zip(ps, range(ngenomes))
results.sort(key=lambda (x): x[0], reverse=True)
for i in results:
p = i[1]
hetcom = notes[p]['hetcom']
hetrare = notes[p]['hetrare']
print 'Patient %d has p=%f homhit=%d hommiss=%d het common=%d rare=%d p=%f' % (
p, i[0], notes[p]['homohit'], notes[p]['homomiss'], hetcom, hetrare,
binom.cdf(min(hetcom, hetrare), hetcom + hetrare, 0.5))
if i[0] < .1:
break
ourvals[words[2]] = words[3]
ourps[words[2]] = float(words[4])
for line in file('/windows/ALL.chr6.ours.vcf'):
words=line.rstrip().split('\t')
snp = words[2]
if snp not in ourvals:
continue
meaning=words[3:5]
if len(meaning[1])>1:
meaning=[meaning[0]]+meaning[1].split(',')
for i in range(ngenomes):
ref=words[i+9]
try:
ref=meaning[int(ref[0])]+meaning[int(ref[2])]
except IndexError as e:
print ref
print meaning
throw(e)
ps[i]=update(ourvals[snp], ref, ps[i], ourps[snp], notes[i])
results = zip(ps, range(ngenomes))
results.sort(key=lambda(x):x[0],reverse=True)
for i in results:
p=i[1]
hetcom=notes[p]['hetcom']
hetrare=notes[p]['hetrare']
print 'Patient %d has p=%f homhit=%d hommiss=%d het common=%d rare=%d p=%f' % (p,i[0],notes[p]['homohit'],notes[p]['homomiss'], hetcom, hetrare, binom.cdf(min(hetcom, hetrare), hetcom+hetrare, 0.5))
if i[0]<.1:
break