def adaptive_parseline(line, index2col):
items = line.strip("\n").split("\t")
if len(items) != len(index2col):
sys.stderr.write(
"Incosistent number of columns between the following\
line and the header line, skipped it:\n\
Line:\n%s\n"
% line
)
return None
col2val = {}
valid_cols = adaptive_columns()
for i, col in index2col.iteritems():
if col in valid_cols:
col2val[col] = items[i].replace("/", ", ")
# Return None if line does not have minimum required fields.
required_cols = [
"count",
"frequencyCount",
"nucleotide",
"vGeneName",
"jGeneName",
"vGeneNameTies",
"jGeneNameTies",
]
for c in required_cols:
if c not in col2val: # or col2val[c] in ["(undefined)", ""]:
return None
count = int(col2val["count"])
freq = float(col2val["frequencyCount"]) / 100.0 # convert to non percentage
nuc = col2val["nucleotide"]
vgene = col2val["vGeneName"]
if vgene == "unresolved":
vgenes = col2val["vGeneNameTies"].split(",")
else:
vgenes = [vgene]
jgene = col2val["jGeneName"]
if jgene == "unresolved":
jgenes = col2val["jGeneNameTies"].split(",")
else:
jgenes = [jgene]
# Clone with required fields
clone = Clone(count, freq, nuc, vgenes, jgenes)
# Additional information if available
# Gene info:
if "dGeneName" in col2val:
dgenestr = col2val["dGeneName"]
if dgenestr == "unresolved":
clone.dgenes = col2val["dGeneNameTies"].split(",")
else:
clone.dgenes = [dgenestr]
if "sequenceStatus" in col2val:
status = col2val["sequenceStatus"].lower()
if status is not None and status == "in":
clone.productive = True
elif status == "out" or status == "stop":
clone.productive = False
else:
sys.stderr.write("Unknown status: %s\n" % status)
if "aminoAcid" in col2val:
clone.cdr3aa = col2val["aminoAcid"]
# Junctional info:
offset = 0
if "vIndex" in col2val:
vindex = int(col2val["vIndex"])
if clone.productive:
# Make sure nuc is inframe:
offset = vindex % 3
nuclen = len(clone.nuc)
endoffset = (nuclen - offset) % 3
clone.nuc = clone.nuc[offset : nuclen - endoffset]
clone.aa = libcommon.nt2aa(clone.nuc)
if clone.cdr3aa:
cdr3len = len(clone.cdr3aa) * 3
endindex = max(len(clone.nuc), vindex + cdr3len)
clone.cdr3nuc = clone.nuc[vindex:endindex]
if "dIndex" in col2val:
clone.firstdpos = int(col2val["dIndex"]) - offset
if "n2Index" in col2val:
n2index = int(col2val["n2Index"])
if n2index != -1:
clone.lastvpos = n2index - 1 - offset
elif clone.firstdpos: # No d5ins
clone.lastvpos = clone.firstdpos - 1
if "jIndex" in col2val:
clone.firstjpos = int(col2val["jIndex"]) - offset
if "n1Index" in col2val:
n1index = int(col2val["n1Index"])
if n1index != -1:
clone.lastdpos = n1index - 1 - offset
elif clone.firstjpos: # No d3ins
clone.lastdpos = clone.firstjpos - 1
# Deletion info:
if "vDeletion" in col2val:
clone.vdel = int(col2val["vDeletion"])
if "d5Deletion" in col2val:
clone.d5del = int(col2val["d5Deletion"])
if "d3Deletion" in col2val:
clone.d3del = int(col2val["d3Deletion"])
if "jDeletion" in col2val:
clone.jdel = int(col2val["jDeletion"])
return clone
def sequenta_parseline(line, index2col):
items = line.strip("\n").split("\t")
if len(items) != len(index2col):
sys.stderr.write(
"Incosistent number of columns between the following\
line and the header line, skipped it:\n\
Line:\n%s\n"
% line
)
return None
col2val = {}
valid_cols = sequenta_columns()
for i, col in index2col.iteritems():
if col in valid_cols:
col2val[col] = items[i]
# Return None if clone is "Water"
if "Patient" in col2val and col2val["Patient"] == "Water":
return None
# Return None if line does not have minimum required fields.
required_cols = [
"Total_Read_Count",
"Log10_Frequency",
"Clone_Sequence",
"V_Segment_Major_Gene",
"J_Segment_Major_Gene",
]
for c in required_cols:
if c not in col2val or col2val[c] in ["NAN", "", "-"]:
return None
count = libcommon.soft_int(col2val["Total_Read_Count"])
try:
freq = 10 ** float(col2val["Log10_Frequency"])
except: # Return None if clone does not have a valid frequency
return None
nuc = col2val["Clone_Sequence"]
vgenes = col2val["V_Segment_Major_Gene"].split("; ")
jgenes = col2val["J_Segment_Major_Gene"].split("; ")
# Clone with required fields
clone = Clone(count, freq, nuc, vgenes, jgenes)
# Additional information if available
# Gene info:
if "D_Segment_Major_Allele" in col2val:
dstr = col2val["D_Segment_Major_Allele"]
if dstr not in ["NAN", "", "-"]:
dalleles = dstr.split("; ")
dgenes = []
for d in dalleles:
dgene = d.split("*")[0]
if dgene not in dgenes:
dgenes.append(dgene)
clone.dgenes = dgenes
clone.dalleles = dalleles
if not clone.dgenes: # no dgenes info
jgroups = get_j_groups(clone.jgenes)
if ["1"] == jgroups:
clone.dgenes = ["TRBD1"]
else:
clone.dgenes = [random.choice(["TRBD1", "TRBD2"])]
if "V_Segment_Major_Allele" in col2val:
clone.valleles = col2val["V_Segment_Major_Allele"].split("; ")
if "J_Segment_Major_Allele" in col2val:
clone.jalleles = col2val["J_Segment_Major_Allele"].split("; ")
# Sequence ID, status and cdr3aa:
if "Sample" in col2val:
clone.samplename = col2val["Sample"]
if "Patient" in col2val:
clone.patient = col2val["Patient"]
if "Clone_Index" in col2val:
clone.id = col2val["Clone_Index"]
if "Is_Good_Frame" in col2val:
if col2val["Is_Good_Frame"].lower() == "true":
clone.productive = True
else:
clone.productive = False
if "Clone_Protein_Sequence" in col2val:
clone.aa = col2val["Clone_Protein_Sequence"].replace("*", "Z")
offset = 0
if "CDR3_Sense_Sequence" in col2val:
clone.cdr3nuc = col2val["CDR3_Sense_Sequence"]
if not re.search(clone.cdr3nuc, clone.nuc):
clone.nuc = libcommon.rc(clone.nuc)
try:
cdr3aa = sequenta_getaa(clone.cdr3nuc)
clone.cdr3aa = cdr3aa
except: # return None if cannot translate cdr3nuc
return None
# Make sure nuc is in frame
cdr3start = re.search(clone.cdr3nuc, clone.nuc).start()
offset = cdr3start % 3
nuclen = len(clone.nuc)
endoffset = (nuclen - offset) % 3
clone.nuc = clone.nuc[offset : nuclen - endoffset]
# Junctional info:
if "V_Segment_Extension_Length" in col2val:
vins = libcommon.soft_int(col2val["V_Segment_Extension_Length"])
clone.lastvpos = vins - 1 - offset
if "N_Bases_adjacent_V" in col2val:
d5ins = col2val["N_Bases_adjacent_V"]
if not d5ins.startswith("-") and d5ins not in ["", "NAN"]:
clone.firstdpos = clone.lastvpos + int(d5ins) + 1
if "J_Segment_Extension_Length" in col2val:
jins = libcommon.soft_int(col2val["J_Segment_Extension_Length"])
clone.firstjpos = len(clone.nuc) - jins
if "N_Bases_adjacent_J" in col2val:
d3ins = col2val["N_Bases_adjacent_J"]
if not d3ins.startswith("-") and d3ins not in ["", "NAN"]:
clone.lastdpos = clone.firstjpos - int(d3ins) - 1
# Deletions:
if "V_Segment_Deletion_Length" in col2val:
vdel = col2val["V_Segment_Deletion_Length"]
if not vdel.startswith("-") and vdel not in ["", "NAN"]:
clone.vdel = libcommon.soft_int(vdel)
if "J_Segment_Deletion_Length" in col2val:
jdel = col2val["J_Segment_Deletion_Length"]
if not jdel.startswith("-") and jdel not in ["", "NAN"]:
clone.jdel = libcommon.soft_int(jdel)
# Special treatment for D info:
d2fulllen = {"TRBD1": 12, "TRBD2": 16}
if "D_Segment_length" in col2val:
dgene = clone.dgenes[0]
dfulllen = d2fulllen[dgene]
dlen = col2val["D_Segment_length"]
if not dlen.startswith("-") and dlen not in ["", "NAN"]:
ddel = dfulllen - int(dlen)
clone.d5del, clone.d3del = get_ddels(ddel)
# clone.d5del = ddel / 2.0
# clone.d3del = ddel - clone.d5del
else: # all D was deleted
clone.d5del, clone.d3del = get_ddels(dfulllen)
# clone.d5del = dfulllen / 2
# clone.d3del = dfulllen - clone.d5del
ndn = clone.firstjpos - clone.lastvpos
clone.firstdpos = clone.lastvpos + ndn / 2 + 1
clone.lastdpos = clone.firstdpos - 1
return clone
def mitcr_parseline(line, index2col):
items = line.strip().split('\t')
if len(items) != len(index2col):
sys.stderr.write("Inconsistent number of columns between the following\
line and the header line, skipped it:\n\
Line:\n%s\n" % line)
return None
col2val = {}
valid_cols = mitcr_columns()
for i, col in index2col.iteritems():
if col in valid_cols:
col2val[col] = items[i]
# Return None if line does not have minimum required fields.
required_cols = ["Read count", "Percentage", "CDR3 nucleotide sequence",
"V segments", "J segments"]
for c in required_cols:
if c not in col2val or not col2val[c]:
return None
count = int(col2val['Read count'])
freq = float(col2val['Percentage'])/100.0
nuc = col2val['CDR3 nucleotide sequence']
vgenes = col2val['V segments'].split(', ')
jgenes = col2val['J segments'].split(', ')
clone = Clone(count, freq, nuc, vgenes, jgenes, cdr3nuc=nuc)
clone.productive = True # Assuming MiTCR only output productive clones
if 'D segments' in col2val:
clone.dgenes = col2val['D segments'].split(', ')
if 'V alleles' in col2val:
clone.valleles = col2val['V alleles'].split(', ')
if 'J alleles' in col2val:
clone.jalleles = col2val['J alleles'].split(', ')
if 'D alleles' in col2val:
clone.dalleles = col2val['D alleles'].split(', ')
if 'CDR3 amino acid sequence' in col2val:
clone.aa = col2val['CDR3 amino acid sequence']
clone.cdr3aa = col2val['CDR3 amino acid sequence']
if 'Last V nucleotide position' in col2val:
clone.lastvpos = int(col2val['Last V nucleotide position'])
if 'First D nucleotide position' in col2val:
clone.firstdpos = int(col2val['First D nucleotide position'])
if 'Last D nucleotide position' in col2val:
clone.lastdpos = int(col2val['Last D nucleotide position'])
if 'First J nucleotide position' in col2val:
clone.firstjpos = int(col2val['First J nucleotide position'])
return clone
