def getJsnps(seq, genes, gene2pos):
pos2snp = {}
jfrag = seq.nuc[seq.jindex:].upper()
jseq = genes[seq.jgene].upper()
#Find where the jfrag is in the jseq:
matches = re.finditer(jfrag, jseq)
starts = [ m.start() for m in matches ]
okstarts = []
pos = gene2pos[seq.jgene] + 3
for s in starts:
e = s + len(jfrag)
if s <= 15 and e >= 10 and e < pos: #the primer should be somewhere to the left (5'end) of the J gene and should cover somewhere after the 10 leftmost nucleotides
extraseq = jseq[e :pos]
cdr3len = len(seq.nuc) - seq.vindex + len(extraseq)
if cdr3len%3 == 0: #inframe
#Translate to cdr3:
newNuc = seq.nuc + extraseq
newaa = iseqlib.nt2aa( newNuc[seq.vindex:] )
if newaa[ len(newaa) -1 ] == 'F' and '*' not in newaa:
if seq.aa not in newaa:
sys.stderr.write('Warning: the new infered aa doesnot contain current aa\n')
okstarts.append(s)
if len(okstarts) > 1:
sys.stderr.write("Looking for SNPs if any on the J fragment. However, mapped to multiple Js. Sequence: %s, %s, %s\n" %(seq.id, seq.nuc, seq.aa))
sys.exit(1)
startPos = starts[0]
endPos = min( [ startPos + len(jfrag), gene2pos[seq.jgene] + 3 ] )
if endPos > len(jseq):
sys.stderr.write("Clone goes beyond downstream of J gene\n")
refjfrag = jseq[startPos:endPos]
seq.jrefstart = startPos
return getSnps(jfrag, refjfrag, startPos)
def fillInSeq(seq, genes, gene2pos):
'''The earlier version of adaptiveTCR tsv files did not always have the complete CDR3 sequences (because reads only support partial CDR3)
This function looked up the original J gene sequences that the clone to mapped to, and fill in the its sequences to have complete CDR3.
'''
jfrag = seq.nuc[seq.jindex:].upper()
jseq = genes[seq.jgene].upper()
pos = gene2pos[seq.jgene] + 3
#find where the jfrag is in the jseq:
matches = re.finditer(jfrag, jseq)
starts = [ m.start() for m in matches ]
#Now finding all possible matches:
fullseqs = []
newcdr3s = []
for s in starts:
e = s + len(jfrag)
if s <= 15 and e >= 10 and e < pos: #the primer should be somewhere to the left (5'end) of the J gene and should cover somewhere after the 10 leftmost nucleotides
extraseq = jseq[e :pos]
cdr3len = len(seq.nuc) - seq.vindex + len(extraseq)
if cdr3len%3 == 0: #inframe
#Translate to cdr3:
newNuc = seq.nuc + extraseq
newaa = iseqlib.nt2aa( newNuc[seq.vindex:] )
if newaa[ len(newaa) -1 ] == 'F' and '*' not in newaa:
if seq.aa not in newaa:
sys.stderr.write('Warning: the new infered aa doesnot contain current aa\n')
fullseqs.append( newNuc )
newcdr3s.append( newaa )
if len(fullseqs) == 0:
sys.stderr.write('Attempted to fill in the right side of the nuc and CDR3 sequences. Zero productive matches found. Sequence looked at was: %s, %s, %s\n' %(seq.id, seq.nuc, seq.aa))
sys.stderr.write("jfrag: %s, jseq: %s, matchesStarts: *%s*\n" %(jfrag, jseq, ','.join(starts)) )
return -1
elif len(fullseqs) > 1:
sys.stderr.write('Attempted to fill in the right side of the nuc and CDR3 sequences. Multiple productive matches found - could not decide. Sequence looked at was: %s, %s, %s\n' %(seq.id, seq.nuc, seq.aa))
return -1
else:
seq.nuc = fullseqs[0]
seq.aa = newcdr3s[0]
seq.jindex = seq.vindex + len(seq.aa)*3
seq.cdr3nuc = seq.nuc[seq.vindex:seq.jindex]
seq.inframenuc = seq.nuc[seq.vindex%3: len(seq.nuc) - ((len(seq.nuc) - seq.vindex%3)%3)]
seq.longaa = iseqlib.nt2aa( seq.inframenuc )
return 1
def __init__(self, name, seq, count, vs, js, freq):
self.name = name
self.seq = seq
self.count = count
self.freq = freq
self.vs = sorted(vs)
self.js = sorted(js)
vstr = ','.join(vs)
jstr = ','.join(js)
self.aa = iseqlib.nt2aa(self.seq)
#self.header = '|'.join([seq, vstr, jstr]) #header example: CASSLRRGGKPGELFF|TRBV5-4|TRBJ2-2
self.header = '|'.join([self.aa, vstr, jstr]) #header example: CASSLRRGGKPGELFF|TRBV5-4|TRBJ2-2
def __init__(self, name, seq, count, vs, js, freq, translate):
self.name = name
self.samples = name.split(',')
self.seq = seq
self.count = count
self.freq = freq
self.vs = sorted(vs)
self.js = sorted(js)
#vstr = ','.join(vs)
#jstr = ','.join(js)
vfams = getGeneFamilies(self.vs)
#jfams = getGeneFamilies(self.js)
vstr = ','.join(vfams)
jstr = ','.join(js)
if translate:
seq = iseqlib.nt2aa(seq)
self.header = '|'.join([seq, vstr, jstr]) #header example: CASSLRRGGKPGELFF|TRBV5|TRBJ2-2 (aminoacid|vfamilies|jgenes)
def __init__(self, line):
items = line.strip().split('\t')
if len(items) < 27:
sys.stderr.write('Wrong tsv format. Expected 27 fields, only have %d\n%s\n' %(len(items), line))
sys.exit(1)
self.id = items[0]
self.nuc = items[2]
self.aa = items[3]
self.normFreq = -1.0
self.normCount = -1
if items[4] != '':
self.normFreq = float(items[4])
if items[5] != '':
self.normCount = int(items[5])
self.freq = float(items[6])
self.count = int(items[7])
self.cdr3len = int(items[8])
self.vfam = items[9]
self.vgene = items[10]
self.vties = items[11]
vgenes = self.vgene.split('/')
if len(vgenes) > 1:
self.vgene = vgenes[0]
vs = self.vties.split(', ')
for v in vgenes[1:]:
if v not in vs:
vs.append(v)
self.vties = ', '.join(vs)
self.vgene = self.vgene.split('*')[0]
self.dgene = items[12]
self.jgene = items[13]
self.jties = items[14]
jgenes = self.jgene.split('/')
if len(jgenes) > 1:
self.jgene = jgenes[0]
js = self.jties.split(', ')
for j in jgenes[1:]:
if j not in js:
js.append(j)
self.jties = ', '.join(js)
self.jgene = self.jgene.split('*')[0]
self.vdel = int(items[15])
self.d5del = int(items[16])
self.d3del = int(items[17])
self.jdel = int(items[18])
self.n2ins = int(items[19])
self.n1ins = int(items[20])
self.status = items[21]
self.vindex = int(items[22])
self.n1index = int(items[23])
self.n2index = int(items[24])
self.dindex = int(items[25])
self.jindex = int(items[26])
self.cdr3nuc = self.nuc[self.vindex:self.jindex]
self.inframenuc = self.nuc[ self.vindex%3: len(self.nuc) - ((len(self.nuc) - self.vindex%3)%3) ]
self.longaa = iseqlib.nt2aa( self.inframenuc )
self.vpos2snp = None
self.jpos2snp = None
