def checkRef(name):
reffile = name +'RefT'
altfile = name + 'AltT'
hg19 = worldbase.Bio.Seq.Genome.HUMAN.hg19(download = True)
print "Loaded hg19"
ref = glob.json(reffile)
print "Loaded Ref"
alt = glob.json(altfile)
print "Loaded Alt"
flip = []
errors = []
keys = ref.keys()
for snppos in keys:
print snppos + '\t' + name
t = snppos.split('pos')
hg19snp = str(hg19[t[0]][int(t[1])-1]).upper()
refsnp = ref[snppos].upper()
altsnp = alt[snppos].upper()
if hg19snp == refsnp:
continue
elif hg19snp == altsnp:
flip.append(snppos)
else:
print "Error: Neither Ref nor Alt of SNP corresponds to hg19 sequence"
errors.append(snppos)
glob.dump(flip, name+'flips')
glob.dump(errors, name+'errors')
return [flip, errors]
def parse1KGvcf(vcffile, outputname):
file = open(vcffile)
outputfile = open(outputname+'Geno', 'w')
ref = {}
alt = {}
lines = file.readlines(1000000)
while(lines != []):
for l in lines:
if l.startswith('#CHROM'):
g = reduce(lambda x,y: x+','+y, l.strip('\n').split('\t')[9:])
outputfile.write('\t'+str(g) +'\n')
if not l.startswith('#'):
tokens = l.strip('\n').split('\t')
f = filter(lambda x: 'GP' in x, tokens[7].split(';'))
if f != []:
pos = 'chr'+f[0].split('=')[1].split(':')[0]+'pos'+f[0].split('=')[1].split(':')[1]
ref[pos] = tokens[3]
alt[pos] = tokens[4]
m=pos +'\t'
for t in tokens[9:]:
m = m + str(int(t[0]) + int(t[2])) + ','
outputfile.write(m.strip(',')+'\n')
lines = file.readlines(1000000)
glob.dump(ref, outputname+'Ref')
glob.dump(alt, outputname+'Alt')
def getsnpgenos(genos, filestruc, chosenSNPs, incarray = 0):
lines = filestruc.genofile.readlines()
snppos = map(lambda x: x.split('\t')[0], lines[1:])
inboth = set(snppos) & set(chosenSNPs)
notingeno = set(filter(lambda x: x not in inboth, chosenSNPs))
try:
genos['lines'] = genos['lines'] +lines[0].split('\t')[1].split(',')
except KeyError:
genos['lines'] = lines[0].split('\t')[1].split(',')
print("Number of Lines in Genotype:" + str(len(lines)))
for l in lines[1:]:
t = l.split('\t')
snp = t[0]
if snp in inboth:
try:
genos[snp] = genos[snp].strip(',') + ','+t[1].strip('\n')
except KeyError:
genos[snp] = t[1].strip('\n')
if incarray == 0:
for s in notingeno:
try:
genos[s] = genos[s].strip(',') + ','+('0,' * filestruc.ln)
except KeyError:
genos[s] = '0,' * filestruc.ln
glob.dump(genos, 'tempgenos')
return genos
def corrRef(flip, name): reffile = name +'RefT' altfile = name + 'AltT' for snp in flip: t = ref[snp] ref[snp] = alt[snp] alt[snp] = t glob.dump(ref, reffile+'flipped') glob.dump(alt, altfile+'flipped')
def filterzeros(arrayname): """take out those that are 0 """ freq = glob.json(arrayname+'freq') for snp in freq.keys(): if freq[snp] == 0 or math.isnan(freq[snp]): del freq[snp] glob.dump(freq, arrayname+'freq')
def makerhash():
rsid2pos = {}
hapmaprsidfile = 'hapmap_rsid_hash_lines'
assert hapmaprsidfile in os.listdir('./'), "Need this file {0}".format(hapmaprsidfile)
file = open('./hapmap_rsid_hash_lines')
with open('./hapmap_rsid_hash_lines') as f:
rlines = f.readlines()
for r in rlines:
t = r.split('\t')
rsid2pos[t[0]] = t[1].strip('\n')
glob.dump(rsid2pos, 'rsid2poshash')
def flipArray(arrayname, flip, error): """flip array snp frequencies (hash) 1-freq for those in snp list inputed as flip input is constructed in the original getarraysnps() function """ try: arrayfreq = glob.json(arrayname+'freq') except: "No array snp frequency file" for snp in flip: arrayfreq[snp] = 1 - arrayfreq[snp] for snp in error: del arrayfreq[snp] glob.dump(arrayfreq, arrayname+'freq')
def parsehapmap():
import parsehapmapgenotypes
ref = {}
alt = {}
genotype = open('hapmapGeno','w')
for c in range(1,23):
"here"
[r,a] = parsehapmapgenotypes.parsehapmapchrom(c)
ref.update(r)
alt.update(a)
with open('../genotypes/hapmapchr'+str(c)+'genotype') as g:
lines = g.readlines()
map(lambda l: genotype.write(l), lines)
genotype.close()
glob.dump(ref, '../genotypes/hapmapRef')
glob.dump(alt, '../genotypes/hapmapAlt')
def filterSNPs(name):
reffile = name +'Ref'
altfile = name + 'Alt'
[ref, alt] = map(lambda x: glob.json(x, ''), [reffile, altfile])
print "Loaded Ref {0}, Alt {1}".format(len(ref),len(alt))
keys = ref.keys()
complsnps = []
for snppos in keys:
if glob.compl[ref[snppos].upper()] == alt[snppos].upper() or ref[snppos].upper() == alt[snppos].upper():
complsnps.append(snppos)
del ref[snppos]
del alt[snppos]
print len(ref)
print len(alt)
glob.dump(ref, reffile+'T')
glob.dump(alt, altfile+'T')
return complsnps
def combinegenos(names, chosenSNPs, out = 'combGenosfile', incarray = 0):
genos = {}
if type(names) is str:
f = Genotypes(names)
genos = getsnpgenos(genos, f, chosenSNPs, incarray)
if type(names) is list:
files = map(lambda x: Genotypes(x), names)
genos = reduce(lambda x,y: getsnpgenos(x, y, chosenSNPs, incarray), [genos]+files)
glob.dump(genos, out+'.json')
genos['lines'] = map(lambda x: x.strip('\n'), genos['lines'])
output = open(out, 'w')
linenames = reduce(lambda x,y: x +',' + y, genos['lines'])
output.write('\t' + linenames +'\n')
for g in genos.keys():
if g != 'lines':
output.write(g + '\t' + genos[g].strip(',') + '\n')
def getarraysnps(report):
print report
file = open(report)
lines = file.readlines()
file.close()
#header = "SNP Name,Sample ID,Allele1 - Top,Allele2 - Top,GC Score,Allele1 - Plus,Allele2 - Plus,Chr,Position,SNP,Theta,R,X,Y,X Raw,Y Raw,B Allele Freq"
#header = "SNP Name,Sample ID,Allele1 - Top,Allele2 - Top,GC Score,Allele1 - Forward,Allele2 - Forward,Allele1 - Plus,Allele2 - Plus,Chr,Position,GT Score,Cluster Sep,SNP,X,Y,X Raw,Y Raw,B Allele,Freq,Log R Ratio,CNV Value,CNV Confidence"
#h = header.split(',')
#h = ['SNP Name', 'Sample ID', 'Allele1 - Top', 'Allele2 - Top', 'GC Score', 'Allele1 - Forward', 'Allele2 - Forward', 'Allele1 - Plus', 'Allele2 - Plus', 'Chr', 'Position', 'GT Score', 'Cluster Sep', 'SNP', 'X', 'Y', 'X Raw', 'Y Raw', 'B Allele Freq', 'Log R Ratio', 'CNV Value', 'CNV Confidence', 'Top Genomic Sequence', 'Plus/Minus Strand', 'Theta', 'R\r\n']
h = 'SNP Name\tSample ID\tAllele1 - Top\tAllele2 - Top\tGC Score\tSNP Index\tAllele1 - Forward\tAllele2 - Forward\tAllele1 - AB\tAllele2 - AB\tAllele1 - Plus\tAllele2 - Plus\tChr\tPosition\tSNP\tILMN Strand\tTop Genomic Sequence\tPlus/Minus Strand\tTheta\tR\tX\tY\tX Raw\tY Raw\tB Allele Freq\r\n'.split('\t')
snpi = h.index("SNP")
chri = h.index("Chr")
posi = h.index("Position")
Yi = h.index("Y")
Xi = h.index("X")
snplist = []
ref = {}
alt = {}
freq = {}
for l in lines:
t = l.split('\t')
try:
if t[chri] not in map(lambda x: str(x), range(1,23)):
continue
else:
snppos = 'chr'+t[chri]+'pos'+t[posi]
snplist.append(snppos)
ref[snppos] = t[snpi].split('/')[0][1]
alt[snppos] = t[snpi].split('/')[1][0]
freq[snppos] = float(t[Yi])/(float(t[Yi])+float(t[Xi]))
except:
continue
glob.dump(snplist, report+'snps')
glob.dump(ref, report+'RefT')
glob.dump(alt, report+'AltT')
glob.dump(freq, report+'freq')