def main(args):
SNV = {}
for line in open(args.SNVvcf):
if line[0] == "#":
continue
content = line.split("\t")
content[1] = int(content[1])
if not content[0] in SNV:
SNV[content[0]] = {}
if not content[1] in SNV[content[0]]:
SNV[content[0]][content[1]] = []
SNV[content[0]][content[1]].append({
"ref": content[3],
"alt": content[4]
})
for line in open(args.SVvcf):
if line[0] == "#" and line[1] == "#":
print line.strip()
continue
elif line[0] == "#":
print "##INFO=<ID=COMPOUND,Number=1,Type=String,Description=\"prints all snps and snvs found within the SV in the following format chr|pos:ref-alt|pos2:ref2->alt2..\">"
print line.strip()
continue
content = line.strip().split("\t")
snp_tag = ";COMPOUND="
chrA, posA, chrB, posB, event_type, INFO, FORMAT = readVCF.readVCFLine(
line)
if not chrA in SNV or not chrB in SNV:
print line.strip()
continue
if chrA == chrB:
snp_tag += retrieve_snps(chrA, posA - args.padding,
posB + args.padding, SNV)
else:
snp_tag += retrieve_snps(chrA, posA - args.padding,
posA + args.padding, SNV)
snp_tag += "|" + retrieve_snps(chrB, posB - args.padding,
posB + args.padding, SNV)
if not snp_tag == ";COMPOUND={}".format(
chrA) and not snp_tag == ";COMPOUND={}|{}".format(chrA, chrB):
content[7] += snp_tag
print "\t".join(content)
def readVCFFile(toRead, chromosomes):
vcfFileName = toRead
vcfInfoLines = []
with open(toRead, 'r') as vcf:
#The first lines should be a number of meta-information lines, prepended by ##.
#Should begin with fileformat. Store these. Check first line for correct format.
line = vcf.readline()
if (not line.startswith("##fileformat=")):
print("VCF file is not in correct format")
return None
else:
print("VCF file seems ok, continuing read")
while (line.startswith("##")):
vcfInfoLines.append(line)
line = vcf.readline()
#A header line prepended by # should follow containing 8 fields, tab-delimited.
#These are in order CHROM, POS, ID, REF, ALT, QUAL, FILTER, INFO. Store in info line list.
fields = line.split('\t')
if (not (line.startswith('#') or len(fields) == 8) ):
print("Header columns missing in VCF file")
return None
else:
vcfInfoLines.append(line)
#All following lines are tab-delmited data lines.
#Store variant data in chromosome item corresponding to CHROM field
numvars = 0
for line in vcf:
numvars += 1
#Feed every data line into readVCF, returning: (chrA,posA,chrB,posB,event_type,description,format)
(chrA,posA,chrB,posB,event_type,description,format) = readVCF.readVCFLine(line)
#Iterate through chromosome list to find match to insert data into
for chromo in chromosomes:
if chromo.name == chrA:
chromo.addVariant(chrA,posA,chrB,posB,event_type,description,format)
break
return (chromosomes,vcfInfoLines)
help="frequency threshold, more common variants are not printed")
parser.add_argument(
'--length',
type=int,
default=10000,
help="length threshold, smaller intergenic variants are not printed")
args, unknown = parser.parse_known_args()
variant_list = []
output = "\"{chrA}\",{chrB},\"{posA}\",\"{posB}\",\"{len}\",\"{var}\",\"{frequency}\",\"{genes}"
sample_ids = readVCF.get_sample_ids(args.vcf)
for line in open(args.vcf):
if not "#" == line[0]:
chrA, posA, chrB, posB, event_type, INFO, format = readVCF.readVCFLine(
line)
# print((chrA, posA, chrB, posB,event_type,INFO,format))
content = line.strip().split("\t")
#have a look in the snpeff field
eff = True
SNPEFF = ""
try:
SNPEFF = content[7].split("EFF=")[1]
except:
if ";ANN=" in content[7]:
SNPEFF = content[7].split("ANN=")[1]
elif ";CSQ=" in content[7]:
eff = False
SNPEFF = content[7].split("CSQ=")[1]
effects = SNPEFF.split(",")
def main(args):
#start by loading the variations
variations = args.variations
ratio = args.overlap
queries = []
with open(variations) as fin:
noOCCTag = 1
infoFound = 0
outputSource = "Not_specified"
for line in fin:
#process and output the metadata
if line.startswith("#") or line.startswith("="):
#find the output source(cnvnator or Findtranslocations)
meta_line = line.replace("#", "")
content = meta_line.split("=")
if (content[0] == "source"):
outputSource = content[1].rstrip().split()[0]
lookForFilter = meta_line.split("=")
#the last infotag will be the Feature tag
if (lookForFilter[0] != "INFO" and noOCCTag
and infoFound == 1):
sys.stdout.write(
"##INFO=<ID=OCC,Number=1,Type=Integer,Description=\"The number of occurances of the event in the database\">\n"
)
sys.stdout.write(
"##INFO=<ID=FRQ,Number=1,Type=Integer,Description=\"The frequency of the event in the database\">\n"
)
sys.stdout.write(line)
infoFound = 0
noFeatureTag = 0
elif (lookForFilter[0] == "INFO"):
sys.stdout.write(line)
infoFound = 1
#there should only be one feature tag per vf file
if (line ==
"INFO=<ID=OCC,Number=1,Type=Integer,Description=\"The number of occurances of the event in the database\">"
):
noOCCTag = 0
else:
sys.stdout.write(line)
else:
#in this case I need to store a query
chrA, startA, endA, chrB, startB, endB, event_type = readVCF.readVCFLine(
outputSource, line)
current_variation = [
chrA.replace("chr", ""),
int(startA),
int(endA),
chrB.replace("chr", ""),
int(startB),
int(endB), event_type, 0, line
] # plus a counter and the variation
queries.append(current_variation)
# at this point queries contains an entry for each variation
#now query each sample.db present in the given folder and store the occurences
if (args.db):
dataBases = glob.glob("{}/*.db".format(os.path.abspath(args.db)))
else:
dataBases = args.files
for sample_db in dataBases:
allVariations = {}
with open(sample_db) as fDB:
for line in fDB:
db_entry = (line.rstrip().split('\t'))
db_entry[0] = db_entry[0].replace("chr", "")
db_entry[1] = db_entry[1].replace("chr", "")
if db_entry[0] in allVariations:
if db_entry[1] in allVariations[db_entry[0]]:
allVariations[db_entry[0]][db_entry[1]].append([
db_entry[2], db_entry[3], db_entry[4], db_entry[5],
db_entry[6], db_entry[7]
])
else:
allVariations[db_entry[0]][db_entry[1]] = [[
db_entry[2], db_entry[3], db_entry[4], db_entry[5],
db_entry[6], db_entry[7]
]]
else:
allVariations[db_entry[0]] = {}
allVariations[db_entry[0]][db_entry[1]] = [[
db_entry[2], db_entry[3], db_entry[4], db_entry[5],
db_entry[6], db_entry[7]
]]
for query in queries:
hit = isVariationInDB(allVariations, query, ratio, args)
if hit:
query[7] += 1 # found hit
for query in sorted(queries, key=itemgetter(7)):
vcf_entry = query[8].rstrip()
content = vcf_entry.split("\t")
content[7] = "{};OCC={};FRQ={}".format(
content[7], query[7], (query[7] / float(len(dataBases))))
print(("\t").join(content))
def main(args):
#start by loading the variations
variations = args.variations
ratio= args.overlap
queries = []
with open(variations) as fin:
noOCCTag=1;
infoFound=0;
outputSource="Not_specified"
for line in fin:
#process and output the metadata
if line.startswith("#") or line.startswith("="):
#find the output source(cnvnator or Findtranslocations)
meta_line=line.replace("#","");
content=meta_line.split("=");
if(content[0] == "source"):
outputSource=content[1].rstrip().split()[0];
lookForFilter=meta_line.split("=");
#the last infotag will be the Feature tag
if(lookForFilter[0] != "INFO" and noOCCTag and infoFound==1):
sys.stdout.write("##INFO=<ID=OCC,Number=1,Type=Integer,Description=\"The number of occurances of the event in the database\">\n");
sys.stdout.write("##INFO=<ID=FRQ,Number=1,Type=Integer,Description=\"The frequency of the event in the database\">\n");
sys.stdout.write(line);
infoFound=0;noFeatureTag=0;
elif(lookForFilter[0] == "INFO"):
sys.stdout.write(line);
infoFound=1;
#there should only be one feature tag per vf file
if(line == "INFO=<ID=OCC,Number=1,Type=Integer,Description=\"The number of occurances of the event in the database\">"):
noOCCTag=0
else:
sys.stdout.write(line)
else:
#in this case I need to store a query
chrA,startA,endA,chrB,startB,endB,event_type =readVCF.readVCFLine(outputSource, line);
current_variation = [chrA.replace("chr",""), int(startA), int(endA), chrB.replace("chr",""), int(startB), int(endB),event_type, 0, line] # plus a counter and the variation
queries.append(current_variation)
# at this point queries contains an entry for each variation
#now query each sample.db present in the given folder and store the occurences
if(args.db):
dataBases=glob.glob("{}/*.db".format(os.path.abspath(args.db)))
else:
dataBases=args.files
for sample_db in dataBases:
allVariations = {}
with open(sample_db) as fDB:
for line in fDB:
db_entry = (line.rstrip().split('\t'))
db_entry[0] = db_entry[0].replace("chr","")
db_entry[1] = db_entry[1].replace("chr","")
if db_entry[0] in allVariations:
if db_entry[1] in allVariations[db_entry[0]]:
allVariations[db_entry[0]][db_entry[1]].append([db_entry[2], db_entry[3], db_entry[4], db_entry[5], db_entry[6],db_entry[7]])
else:
allVariations[db_entry[0]][db_entry[1]] = [[db_entry[2], db_entry[3], db_entry[4], db_entry[5], db_entry[6],db_entry[7]]]
else:
allVariations[db_entry[0]] = {}
allVariations[db_entry[0]][db_entry[1]] = [[db_entry[2], db_entry[3], db_entry[4], db_entry[5], db_entry[6],db_entry[7]]]
for query in queries:
hit,allVariations = isVariationInDB(allVariations, query,ratio,args)
if hit:
query[7] += 1 # found hit
for query in sorted(queries, key=itemgetter(7)):
vcf_entry = query[8].rstrip()
sys.stdout.write("{};OCC={};FRQ={}\n".format(vcf_entry, query[7],(query[7]/float(len(dataBases))) ) )
def main(args):
bin_size = 0
coverage = {}
#read the coverage file
with open(args.tab) as fin:
for line in fin:
if (line[0] == "#"):
pass
else:
content = line.strip().split("\t")
bin_size = int(content[2]) - int(content[1])
if not content[0].replace("chr", "").replace("CHR",
"") in coverage:
coverage[content[0].replace("chr", "").replace(
"CHR", "")] = [float(content[3])]
else:
coverage[content[0].replace("chr",
"").replace("CHR", "")].append(
float(content[3]))
#compute the coverage across each chromosome
chr_cov = chromosome_coverage(bin_size, coverage)
#use the coverage to genotype the CNVS
RD = 0
PE = 0
SR = 0
GT = 0
format_print = 1
with open(args.vcf) as fin:
for line in fin:
if (line[0] == "#" and line[1] == "#"):
if ("#FORMAT=<ID=RD,Number=1,Type=Float,Description=\"The read depth fraction at the variant\">"
in line):
RD = 1
elif ("#FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of discordant pairs\">"
in line):
PE = 1
elif ("#FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of split reads\">"
):
SR = 1
elif ("#FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">"
):
GT = 1
print(line.strip())
elif (line[0] == "#"):
out_line = line.strip() + "\tFORMAT\t" + args.sample
if format_print:
if not RD:
print(
"##FORMAT=<ID=RD,Number=1,Type=Float,Description=\"The read depth fraction at the variant\">"
)
if not PE:
print(
"##FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of discordant pairs\">"
)
if not SR:
print(
"##FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of split reads\">"
)
if not GT:
print(
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">"
)
print(out_line)
format_print = 0
else:
RD = "."
PE = "0"
SR = 0
GT = "./."
chrA, startA, endA, chrB, startB, endB, event_type = readVCF.readVCFLine(
None, line)
out_line = line.strip()
try:
chrA_cov = chr_cov[chrA]
#calculate the coverage across intra chromosomal variants
if (chrA == chrB):
var_cov = mean_coverage(chrA, startA, endA, bin_size,
coverage)
RD = (var_cov / float(chr_cov[chrA]))
if ("DUP" in event_type):
if (RD < 1.4) * chr_cov[chrA]:
GT = "0/1"
elif ("DEL" in event_type):
if (RD > 0.1 * chr_cov[chrA]):
GT = "0/1"
else:
GT = "1/1"
except:
pass
RD = str(RD)
#compute the number of discordant pairs
if (";LTE=" in line):
LTE = line.split(";LTE=")[1]
LTE = LTE.split(";")[0]
PE = LTE
#generate the format data
format_line = [GT, PE, RD, str(SR)]
out_line = out_line + "\t" + "GT:PE:RD:SR" + "\t" + ":".join(
format_line)
print(out_line)
def main(args):
bin_size=0;
coverage={}
#read the coverage file
with open(args.tab) as fin:
for line in fin:
if(line[0]=="#"):
pass
else:
content=line.strip().split("\t")
bin_size= int(content[2]) - int(content[1])
if not content[0].replace("chr","").replace("CHR","") in coverage:
coverage[content[0].replace("chr","").replace("CHR","")]=[float(content[3])]
else:
coverage[content[0].replace("chr","").replace("CHR","")].append(float(content[3]))
#compute the coverage across each chromosome
chr_cov=chromosome_coverage(bin_size,coverage);
#use the coverage to genotype the CNVS
RD=0;
PE=0;
SR=0;
GT=0;
format_print=1;
with open(args.vcf) as fin:
for line in fin:
if(line[0]=="#" and line[1] == "#" ):
if ("#FORMAT=<ID=RD,Number=1,Type=Float,Description=\"The read depth fraction at the variant\">" in line):
RD = 1;
elif("#FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of discordant pairs\">" in line):
PE = 1;
elif("#FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of split reads\">" ):
SR = 1;
elif("#FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">" ):
GT = 1;
print(line.strip());
elif(line[0] == "#"):
out_line=line.strip()+"\tFORMAT\t"+args.sample
if format_print:
if not RD:
print("##FORMAT=<ID=RD,Number=1,Type=Float,Description=\"The read depth fraction at the variant\">");
if not PE:
print("##FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of discordant pairs\">" );
if not SR:
print("##FORMAT=<ID=PE,Number=1,Type=Integer,Description=\"The number of split reads\">");
if not GT:
print("##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">");
print(out_line)
format_print=0;
else:
RD="."
PE="0";
SR=0;
GT="./."
chrA,startA,endA,chrB,startB,endB,event_type =readVCF.readVCFLine(None, line);
out_line=line.strip();
try:
chrA_cov=chr_cov[chrA];
#calculate the coverage across intra chromosomal variants
if(chrA == chrB):
var_cov=mean_coverage(chrA,startA,endA,bin_size,coverage)
RD=( var_cov/float( chr_cov[chrA] ) )
if("DUP" in event_type):
if(RD < 1.4)*chr_cov[chrA]:
GT="0/1"
elif("DEL" in event_type):
if(RD > 0.1*chr_cov[chrA]):
GT="0/1"
else:
GT="1/1"
except:
pass
RD=str(RD)
#compute the number of discordant pairs
if(";LTE=" in line):
LTE=line.split(";LTE=")[1];
LTE=LTE.split(";")[0]
PE=LTE
#generate the format data
format_line=[GT,PE,RD,str(SR)]
out_line=out_line+"\t"+"GT:PE:RD:SR" + "\t" + ":".join(format_line)
print(out_line)
def main(args):
allVariations = {}
tollerance = args.tollerance
fixed =0
if args.fixed:
fixed =1
for variation_file in [item for sublist in args.variations for item in sublist] :
outputSource=None;
collapsedVariations = {} # this will contain the SVs of this file, but collapsing those that are close
with open(variation_file) as fin:
#memorize all variations
for line in fin:
if line.startswith("#") or line.startswith("="):
#find the output source(cnvnator or Findtranslocations
line=line.replace("#","");
content=line.split("=");
if(content[0] == "source"):
outputSource=content[1].rstrip().split()[0];
continue
chrA,startA,endA,chrB,startB,endB,event_type =readVCF.readVCFLine(outputSource,line);
startA = startA - tollerance
if startA < 0:
startA = 0
startB = startB - tollerance
if startB < 0:
startB = 0
current_variation = [chrA, startA , endA + tollerance, chrB, startB, endB + tollerance,event_type]
collapsedVariations = populate_DB(collapsedVariations, current_variation, True, 0, fixed )
##collapse again in order to avoid problems with areas that have become too close one to onther
elemnets_before = 0
for chrA in collapsedVariations:
for chrB in collapsedVariations[chrA] :
for collapsedVariation in collapsedVariations[chrA][chrB]:
elemnets_before += 1
elemnets_after = 0
while elemnets_before != elemnets_after:
collapsedVariationsFinal = {}
elemnets_before = 0
elemnets_after = 0
for chrA in collapsedVariations:
for chrB in collapsedVariations[chrA] :
for collapsedVariation in collapsedVariations[chrA][chrB]:
current_variation = [chrA, collapsedVariation[0], collapsedVariation[1], chrB, collapsedVariation[2], collapsedVariation[3],collapsedVariation[4]]
collapsedVariationsFinal = populate_DB(collapsedVariationsFinal, current_variation, True, 0,fixed)
elemnets_before += 1
for chrA in collapsedVariationsFinal:
for chrB in collapsedVariationsFinal[chrA] :
for collapsedVariation in collapsedVariationsFinal[chrA][chrB]:
elemnets_after += 1
collapsedVariations.clear()
collapsedVariations = collapsedVariationsFinal
#now populate the DB
for chrA in collapsedVariations:
for chrB in collapsedVariations[chrA] :
for collapsedVariation in collapsedVariations[chrA][chrB]:
current_variation = [chrA, collapsedVariation[0], collapsedVariation[1], chrB, collapsedVariation[2], collapsedVariation[3],collapsedVariation[4]]
allVariations = populate_DB(allVariations, current_variation, False, 0,fixed)
for chrA in allVariations:
for chrB in allVariations[chrA] :
for event in allVariations[chrA][chrB]:
print "{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}".format(chrA, chrB, event[0], event[1], event[2], event[3], event[4], event[5])
def main(args):
#print "memorizing masked reagions"
toBeMaskedElements = {}
featureEntries = []
for bed_file in [item for sublist in args.bed_files for item in sublist]:
toBeMaskedElements[bed_file] = {}
bedlabel = bed_file.split("/")[-1]
bedlabel = bedlabel.split(".")[0]
featureEntries.append(
"##INFO=<ID={0},Number=2,Type=String,Description=\"Genomic features of regions A and B\">\n"
.format(bedlabel))
#sys.stdout.write("memorizing {} ...".format(bed_file))
with open(bed_file) as fin:
rows = (line.rstrip().split('\t') for line in fin)
for row in rows:
if not row[0].startswith("#"):
try:
toBeMaskedElements[bed_file][row[0].replace(
"chr",
"").replace("Chr",
"")].append([row[1], row[2], row[3]])
except:
toBeMaskedElements[bed_file][row[0].replace(
"chr",
"").replace("Chr",
"")] = [[row[1], row[2], row[3]]]
#sys.stdout.write("done\n")
#print "sorting masked regions"
for bedFiles in toBeMaskedElements:
for chr in toBeMaskedElements[bedFiles]:
sorted(toBeMaskedElements[bedFiles][chr], key=itemgetter(1))
#print toBeMaskedElements
#UncompressedVariations = [];
#CompressedVariations = []
#currentVariation = UncompressedVariations[0]
with open(args.variations) as fin:
UncompressedVariations = (line.rstrip().split('\t') for line in fin)
noFeatureTag = 1
infoFound = 0
for row in UncompressedVariations:
#the first charachter is # in the metadata, otherwise it is c,C or a number
if (row[0][0] != "#"):
chr_1, chr_1_start, chr_1_end, chr_2, chr_2_start, chr_2_end, event_type = readVCF.readVCFLine(
outputSource, "\t".join(row))
chr_1_start = int(chr_1_start)
chr_1_end = int(chr_1_end)
outrow = "\t".join(row)
outRow = outrow.replace("\n", "")
sys.stdout.write(outRow)
## Chategorise the two breackpoints
for bedFile in toBeMaskedElements:
bedlabel = bedFile.split("/")[-1]
bedlabel = bedlabel.split(".")[0]
for j in range(2):
if j == 0:
variation_chr = chr_1.replace("chr", "").replace(
"Chr", "")
variation_start = chr_1_start
variation_end = chr_1_end
else:
variation_chr = chr_2.replace("chr", "").replace(
"Chr", "")
variation_start = chr_2_start
variation_end = chr_2_end
stopSearching = 0
if variation_chr in toBeMaskedElements[bedFile]:
categorized = 0
i = 0
max_Size = 0
max_Size_type = ""
while i < len(toBeMaskedElements[bedFile]
[variation_chr]):
repeat_start = int(toBeMaskedElements[bedFile]
[variation_chr][i][0])
repeat_end = int(toBeMaskedElements[bedFile]
[variation_chr][i][1])
repeat_type = toBeMaskedElements[bedFile][
variation_chr][i][2]
i += 1
overlap = 0
if variation_start > repeat_end:
stopSearching = 1
elif variation_start <= repeat_start and variation_end >= repeat_end and variation_start < repeat_end:
overlap = repeat_end - repeat_start + 1
elif variation_start <= repeat_start and variation_end <= repeat_end and variation_start < repeat_end:
overlap = variation_end - repeat_start + 1
elif variation_start >= repeat_start and variation_end >= repeat_end and variation_start < repeat_end:
overlap = repeat_end - variation_start + 1
elif variation_start >= repeat_start and variation_end <= repeat_end and variation_start < repeat_end:
overlap = variation_end - variation_start + 1
if overlap > max_Size:
max_Size = overlap
categorized = 1
max_Size_type = "{}".format(repeat_type)
if categorized == 0:
if (j == 0):
sys.stdout.write(
";{0}=NoCategory".format(bedlabel))
else:
sys.stdout.write(",NoCategory")
else:
if (j == 0):
sys.stdout.write(";{0}={1}".format(
bedlabel, max_Size_type))
else:
sys.stdout.write(
",{}".format(max_Size_type))
else:
if (j == 0):
sys.stdout.write(
";{}=NotPresentInRef".format(bedlabel))
else:
sys.stdout.write(",NotPresentInRef")
sys.stdout.write("\n")
#asasfasfasf
else:
lookForFilter = row[0].split("=")
line = row[0].replace("#", "")
content = line.split("=")
if (content[0] == "source"):
outputSource = content[1].rstrip().split()[0]
#the last infotag will be the Feature tag
if (lookForFilter[0] != "##INFO" and infoFound == 1):
for entry in featureEntries:
sys.stdout.write(entry)
sys.stdout.write(row[0] + "\n")
infoFound = 0
elif (lookForFilter[0] == "##INFO"):
sys.stdout.write(row[0] + "\n")
infoFound = 1
#there should only be one feature tag per vf file
if row[0] in featureEntries: featureEntries.remove(row[0])
else:
sys.stdout.write("\t".join(row) + "\n")
return 0
def main(args):
#print "memorizing masked reagions"
toBeMaskedElements = {};
featureEntries=[];
for bed_file in [item for sublist in args.bed_files for item in sublist] :
toBeMaskedElements[bed_file]={};
bedlabel=bed_file.split("/")[-1]
bedlabel=bedlabel.split(".")[0]
featureEntries.append("##INFO=<ID={0},Number=2,Type=String,Description=\"Genomic features of regions A and B\">\n".format(bedlabel))
#sys.stdout.write("memorizing {} ...".format(bed_file))
with open(bed_file) as fin:
rows = ( line.rstrip().split('\t') for line in fin)
for row in rows:
if not row[0].startswith("#"):
try:
toBeMaskedElements[bed_file][row[0].replace("chr","").replace("Chr","")].append([row[1], row[2], row[3]])
except :
toBeMaskedElements[bed_file][row[0].replace("chr","").replace("Chr","")] = [[row[1], row[2], row[3]]]
#sys.stdout.write("done\n")
#print "sorting masked regions"
for bedFiles in toBeMaskedElements:
for chr in toBeMaskedElements[bedFiles]:
sorted(toBeMaskedElements[bedFiles][chr], key=itemgetter(1))
#print toBeMaskedElements
#UncompressedVariations = [];
#CompressedVariations = []
#currentVariation = UncompressedVariations[0]
with open(args.variations) as fin:
UncompressedVariations = ( line.rstrip().split('\t') for line in fin)
noFeatureTag=1;
infoFound=0;
for row in UncompressedVariations:
#the first charachter is # in the metadata, otherwise it is c,C or a number
if(row[0][0] != "#"):
chr_1,chr_1_start,chr_1_end,chr_2,chr_2_start,chr_2_end,event_type =readVCF.readVCFLine(outputSource,"\t".join(row));
chr_1_start=int(chr_1_start);
chr_1_end=int(chr_1_end);
outrow="\t".join(row);
outRow=outrow.replace("\n","");
sys.stdout.write(outRow)
## Chategorise the two breackpoints
for bedFile in toBeMaskedElements:
bedlabel=bedFile.split("/")[-1]
bedlabel=bedlabel.split(".")[0]
for j in range(2):
if j == 0:
variation_chr = chr_1.replace("chr","").replace("Chr","")
variation_start = chr_1_start
variation_end = chr_1_end
else:
variation_chr = chr_2.replace("chr","").replace("Chr","")
variation_start = chr_2_start
variation_end = chr_2_end
stopSearching = 0
if variation_chr in toBeMaskedElements[bedFile]:
categorized = 0
i = 0
max_Size = 0
max_Size_type = ""
while i < len(toBeMaskedElements[bedFile][variation_chr]) :
repeat_start = int(toBeMaskedElements[bedFile][variation_chr][i][0])
repeat_end = int(toBeMaskedElements[bedFile][variation_chr][i][1])
repeat_type = toBeMaskedElements[bedFile][variation_chr][i][2]
i += 1
overlap = 0
if variation_start > repeat_end:
stopSearching = 1
elif variation_start <= repeat_start and variation_end >= repeat_end and variation_start < repeat_end:
overlap = repeat_end - repeat_start + 1
elif variation_start <= repeat_start and variation_end <= repeat_end and variation_start < repeat_end:
overlap = variation_end - repeat_start + 1
elif variation_start >= repeat_start and variation_end >= repeat_end and variation_start < repeat_end:
overlap = repeat_end - variation_start + 1
elif variation_start >= repeat_start and variation_end <= repeat_end and variation_start < repeat_end:
overlap = variation_end - variation_start + 1
if overlap > max_Size:
max_Size = overlap
categorized = 1
max_Size_type = "{}".format(repeat_type)
if categorized == 0:
if( j == 0):
sys.stdout.write(";{0}=NoCategory".format(bedlabel))
else:
sys.stdout.write(",NoCategory")
else:
if(j == 0):
sys.stdout.write(";{0}={1}".format(bedlabel,max_Size_type))
else:
sys.stdout.write(",{}".format(max_Size_type))
else:
if(j == 0):
sys.stdout.write(";{}=NotPresentInRef".format(bedlabel))
else:
sys.stdout.write(",NotPresentInRef")
sys.stdout.write("\n")
#asasfasfasf
else:
lookForFilter=row[0].split("=");
line=row[0].replace("#","");
content=line.split("=");
if(content[0] == "source"):
outputSource=content[1].rstrip().split()[0];
#the last infotag will be the Feature tag
if(lookForFilter[0] !="##INFO" and infoFound==1):
for entry in featureEntries:
sys.stdout.write(entry);
sys.stdout.write(row[0]+"\n");
infoFound=0;
elif(lookForFilter[0] == "##INFO"):
sys.stdout.write(row[0]+"\n");
infoFound=1;
#there should only be one feature tag per vf file
if row[0] in featureEntries: featureEntries.remove(row[0])
else:
sys.stdout.write("\t".join(row)+"\n");
return 0
def main(Data,GC_hist,args):
if args.vcf:
for line in open(args.vcf):
if line[0] == "#" and line[1] == "#":
print line.strip()
elif line[0] == "#":
print "##INFO=<ID={},Number=1,Type=Float,Description=\"estimated copy number\">".format("AMYCNE")
print "##INFO=<ID={},Number=2,Type=Float,Description=\"99% confidence interval around the estimated CN\">".format("AMYCNECI")
print "##INFO=<ID={},Number=1,Type=Float,Description=\"ratio of bins used for CN estimation\">".format("BIN_RATIO")
print "##INFO=<ID={},Number=1,Type=Float,Description=\"mean coverage of the reference bins\">".format("REFCOV")
print line.strip()
else:
chrA,posA,chrB,posB,event_type,INFO,FORMAT = readVCF.readVCFLine(line)
if chrA == chrB:
if chrA in Data:
cn, gc, length,ref,bins,used_bins,bin_list =common.regional_cn_est( Data ,GC_hist, [chrA,posA,posB],args.Q )
ci="(0,0)"
CNE = round( cn*args.plody )
if CNE < 0:
CNE = -1
else:
for i in range(0,len(bin_list)):
bin_list[i]=bin_list[i]*args.plody
SEM=(numpy.std(bin_list)/numpy.sqrt(used_bins))
ci="{},{}".format(round(cn*args.plody-SEM*3,2),round(cn*args.plody+SEM*3,2))
content=line.split("\t")
if bins == 0:
bins =-1
content[7] += ";AMYCNE=" + str(int(CNE)) + ";BIN_RATIO=" + str(used_bins/float(bins)) + ";REFCOV=" + str(round(ref,2)) + ";AMYCNECI=" + ci
new_line="\t".join(content)
print new_line.strip()
else:
print line.strip()
elif args.bed:
for line in open(args.bed):
if line[0] == "#":
print "{}\t{}\t{}\t{}\t{}".format(line.strip(),"CN","CI","RATIO","Refcov")
continue
content=line.split("\t")
chrA=content[0]
chrB=chrA
posA=content[1]
posB=content[2]
if not chrA in Data and chrA.replace("chr",""):
chrA=chrA.replace("chr","")
if chrA in Data:
cn, gc, length,ref,bins,used_bins,bin_list =common.regional_cn_est( Data ,GC_hist, [chrA,posA,posB],args.Q )
ci="(0,0)"
CNE = int( round( cn*args.plody ) )
if CNE < 0:
CNE = -1
else:
for i in range(0,len(bin_list)):
bin_list[i]=bin_list[i]*args.plody
SEM=(numpy.std(bin_list)/numpy.sqrt(used_bins))
ci="{},{}".format(round(cn*args.plody-SEM*3,2),round(cn*args.plody+SEM*3,2))
content=line.split("\t")
if bins == 0:
bins =-1
print "{}\t{}\t{}\t{}\t{}".format(line.strip(), CNE,ci, used_bins/float(bins), str(round(ref,2)))
else:
print "ERROR: no input source, please select vcf or bed"
def extract_splits(args, ws0):
if args.bed:
input_file = args.bed
elif args.vcf:
input_file = args.vcf
else:
print "error: missing bed or vcf"
quit()
if not args.sample:
args.sample = args.bam.split("/")[-1].split(".")[0]
if not args.working_dir:
args.working_dir = args.bam.split("/")[-1].split(".")[0]
if args.repeatmask:
conn = sqlite3.connect(args.repeatmask)
c = conn.cursor()
row = 1
detected_splits = {}
i = 0
for line in open(input_file):
if line[0] == "#":
continue
if args.bed:
content = line.strip().split()
args.chrA = content[0]
args.posA = int(content[1])
args.chrB = content[2]
args.posB = int(content[3])
args.type = content[4]
args.orientationA = ""
args.orientationB = ""
args.id = str(i)
var_id = str(i)
args.lengthA = ""
args.lengthB = ""
args.regionA = ""
args.regionB = ""
insertion_seq = ""
homology_seq = ""
args.regionAsegments = ()
args.regionBsegments = ()
args.contigSegments = ()
args.HomologySegments = ()
args.repeatA = ""
args.repeatB = ""
i += 1
elif args.vcf:
chrA, posA, chrB, posB, event_type, INFO, FORMAT = readVCF.readVCFLine(
line)
args.chrA = chrA
args.posA = int(posA)
args.chrB = chrB
args.posB = int(posB)
args.type = event_type
args.orientationA = ""
args.orientationB = ""
args.id = line.strip().split("\t")[2]
var_id = line.strip().split("\t")[2]
args.lengthA = ""
args.lengthB = ""
args.regionA = ""
args.regionB = ""
insertion_seq = ""
homology_seq = ""
args.regionAsegments = ()
args.regionBsegments = ()
args.contigSegments = ()
args.HomologySegments = ()
args.repeatA = ""
args.repeatB = ""
i += 1
found = FindTranslocations.main(args)
splits = 0
bp_homology = ""
insertions = ""
deletions = ""
sucess = False
print found
if found:
wd = os.path.join(args.working_dir, var_id)
softclips = os.path.join(wd, "splits.sam")
print "python consensus.py {} {} > {}/consensus.fa".format(
wd, softclips, wd)
os.system("python consensus.py {} {} > {}/consensus.fa".format(
wd, softclips, wd))
for line in open("{}/consensus.fa".format(wd)):
splits = int(line.strip().split()[2])
print splits
break
try:
os.system("bwa mem {} {} > {}".format(
args.fa,
os.path.join(args.working_dir, var_id, "consensus.fa"),
os.path.join(wd, "aligned_consensus.sam")))
args, sucess, contig, bp_homology, homology_seq, insertions, insertion_seq, deletions = retrieve_pos(
args,
os.path.join(args.working_dir, var_id,
"aligned_consensus.sam"))
print success
except:
homology_seq = "WARNING:unable to determine the breakpoint sequence"
else:
wd = os.path.join(args.working_dir, var_id)
os.system("samtools view {} {}:{}-{} > {}/regionA.sam".format(
args.bam, args.chrA, args.posA - args.padding,
args.posA + args.padding, wd))
os.system("samtools view {} {}:{}-{} > {}/regionB.sam".format(
args.bam, args.chrB, args.posB - args.padding,
args.posB + args.padding, wd))
os.system(
"cat {}/regionA.sam {}/regionB.sam > {}/region.sam".format(
wd, wd, wd))
os.system("python bam2fa.py {}/region.sam > {}/region.fq".format(
wd, wd))
trials = [20, 60, 90]
for k in trials:
print "ABYSS -c {} -e {} -k {} -o {}_{}.fa {} > /dev/null 2>&1".format(
1, 10, k, os.path.join(args.working_dir, var_id, "abyss"),
k, os.path.join(wd, "region.fq"))
os.system(
"ABYSS -c {} -e {} -k {} -o {}_{}.fa {} > /dev/null 2>&1".
format(1, 10, k,
os.path.join(args.working_dir, var_id, "abyss"), k,
os.path.join(wd, "region.fq")))
os.system("cat {}_20.fa {}_60.fa {}_90.fa > {}".format(
os.path.join(args.working_dir, var_id, "abyss"),
os.path.join(args.working_dir, var_id, "abyss"),
os.path.join(args.working_dir, var_id, "abyss"),
os.path.join(args.working_dir, var_id, "abyss.fa")))
if not os.stat(os.path.join(args.working_dir, var_id,
"abyss.fa")).st_size == 0:
os.system("bwa mem {} {} > {}".format(
args.fa, os.path.join(args.working_dir, var_id,
"abyss.fa"),
os.path.join(wd, "aligned_contig.sam")))
try:
args, sucess, contig, bp_homology, homology_seq, insertions, insertion_seq, deletions = retrieve_pos(
args, os.path.join(wd, "aligned_contig.sam"))
except:
homology_seq = "WARNING:unable to determine the breakpoint sequence"
if not sucess:
contig = ""
distanceA = ""
distanceB = ""
if args.repeatmask:
distanceA, args.repeatA = find_repeat(args.chrA, args.posA, c)
distanceB, args.repeatB = find_repeat(args.chrB, args.posB, c)
snpDistanceA, snpsA = find_snps(args.chrA, args.posA,
args.snp_distance, args.bam, wd,
args.fa)
snpDistanceB, snpsB = find_snps(args.chrB, args.posB,
args.snp_distance, args.bam, wd,
args.fa)
row_content = [
args.sample, var_id, args.type, splits, args.chrA, args.posA,
args.orientationA, args.repeatA, distanceA, snpsA, snpDistanceA,
args.chrB, args.posB, args.orientationB, args.repeatB, distanceB,
snpsB, snpDistanceB, bp_homology, args.HomologySegments,
insertions, insertion_seq, args.lengthA, args.lengthB,
len(contig), args.regionAsegments, args.regionBsegments,
args.contigSegments
]
j = 0
for item in row_content:
if j in [19, 25, 26, 27]:
ws0.write_rich_text(row, j, item)
else:
ws0.write(row, j, item)
j += 1
row += 1
