def genotyping(read_dir,
reference_type,
region_list,
num_editdist,
nthreads,
max_sample,
assembly,
out_dir,
verbose):
for database_name in region_list:
# Extract variants, backbone sequence, and other sequeces
typing_common.extract_database_if_not_exists(database_name,
[]) # locus_list
# Build HISAT2's graph index
typing_common.build_index_if_not_exists(database_name,
"hisat2",
"graph",
1, # threads
verbose)
if not os.path.exists(read_dir):
print >> sys.stderr, "Error: %s does not exist." % read_dir
sys.exit(1)
if out_dir != "" and not os.path.exists(out_dir):
os.mkdir(out_dir)
# fastq files
fq_fnames = glob.glob("%s/*.extracted.1.fq.gz" % read_dir)
lock = threading.Lock()
threads = []
for t in range(nthreads):
thread = myThread(lock,
fq_fnames,
reference_type,
region_list,
num_editdist,
max_sample,
assembly,
out_dir,
verbose)
thread.start()
threads.append(thread)
for thread in threads:
thread.join()
def build_genotype_genome(base_fname, inter_gap, intra_gap, threads,
database_list, use_clinvar, use_commonvar, aligner,
graph_index, verbose):
# Download HISAT2 index
typing_common.download_genome_and_index()
# Load genomic sequences
chr_dic, chr_names, chr_full_names = typing_common.read_genome("genome.fa")
genotype_vars = {}
genotype_haplotypes = {}
genotype_clnsig = {}
if use_clinvar:
# Extract variants from the ClinVar database
CLINVAR_fnames = [
"clinvar.vcf.gz", "clinvar.snp", "clinvar.haplotype",
"clinvar.clnsig"
]
if not typing_common.check_files(CLINVAR_fnames):
if not os.path.exists("clinvar.vcf.gz"):
os.system("wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/"\
"vcf_GRCh38/archive/2017/clinvar_20170404.vcf.gz")
assert os.path.exists("clinvar.vcf.gz")
extract_cmd = ["hisat2_extract_snps_haplotypes_VCF.py"]
extract_cmd += [
"--inter-gap",
str(inter_gap), "--intra-gap",
str(intra_gap), "--genotype-vcf", "clinvar.vcf.gz",
"genome.fa", "/dev/null", "clinvar"
]
if verbose:
print("\tRunning:", ' '.join(extract_cmd), file=sys.stderr)
proc = subprocess.Popen(extract_cmd,
stdout=open("/dev/null", 'w'),
stderr=open("/dev/null", 'w'))
proc.communicate()
if not typing_common.check_files(CLINVAR_fnames):
print("Error: extract variants from clinvar failed!",
file=sys.stderr)
sys.exit(1)
# Read variants to be genotyped
genotype_vars = typing_common.read_variants("clinvar.snp")
# Read haplotypes
genotype_haplotypes = typing_common.read_haplotypes(
"clinvar.haplotype")
# Read information about clinical significance
genotype_clnsig = read_clnsig("clinvar.clnsig")
if use_commonvar:
# Extract variants from dbSNP database
# TODO: CB Write script to make local uptodate SNP database from dbSNP
# ftp://ftp.ncbi.nlm.nih.gov/snp/database/README.create_local_dbSNP.txt
commonvar_fbase = "snp144Common"
commonvar_fnames = [
"%s.snp" % commonvar_fbase,
"%s.haplotype" % commonvar_fbase
]
if not typing_common.check_files(commonvar_fnames):
if not os.path.exists("%s.txt.gz" % commonvar_fbase):
os.system("wget http://hgdownload.cse.ucsc.edu/goldenPath/hg38/"\
"database/%s.txt.gz" % commonvar_fbase)
assert os.path.exists("%s.txt.gz" % commonvar_fbase)
os.system("gzip -cd %s.txt.gz "\
"| awk 'BEGIN{OFS=\"\t\"} "\
"{if($2 ~ /^chr/) {$2 = substr($2, 4)}; "\
"if($2 == \"M\") {$2 = \"MT\"} print}' > %s.txt" \
% (commonvar_fbase, commonvar_fbase))
extract_cmd = [
"hisat2_extract_snps_haplotypes_UCSC.py", "--inter-gap",
str(inter_gap), "--intra-gap",
str(intra_gap), "genome.fa",
"%s.txt" % commonvar_fbase, commonvar_fbase
]
if verbose:
print("\tRunning:", ' '.join(extract_cmd), file=sys.stderr)
proc = subprocess.Popen(extract_cmd,
stdout=open("/dev/null", 'w'),
stderr=open("/dev/null", 'w'))
proc.communicate()
if not typing_common.check_files(commonvar_fnames):
print("Error: extract variants from clinvar failed!",
file=sys.stderr)
sys.exit(1)
# Read variants to be genotyped
genotype_vars = typing_common.read_variants(commonvar_fnames[0])
# Read haplotypes
genotype_haplotypes = typing_common.read_haplotypes(
commonvar_fnames[1])
# Genes to be genotyped
genotype_genes = {}
# Read genes or genomics regions
for database_name in database_list:
# Extract HLA variants, backbone sequence, and other sequeces
typing_common.extract_database_if_not_exists(
database_name,
[], # locus_list
inter_gap,
intra_gap,
True, # partial?
verbose)
locus_fname = "%s.locus" % database_name
assert os.path.exists(locus_fname)
for line in open(locus_fname):
locus_name, \
chr, \
left, \
right, \
length, \
exon_str, \
strand \
= line.strip().split()
left = int(left)
right = int(right)
length = int(length)
if chr not in chr_names:
continue
if chr not in genotype_genes:
genotype_genes[chr] = []
genotype_genes[chr].append([
left, right, length, locus_name, database_name, exon_str,
strand
])
# Write genotype genome
var_num = 0
haplotype_num = 0
genome_out_file = open("%s.fa" % base_fname, 'w')
locus_out_file = open("%s.locus" % base_fname, 'w')
var_out_file = open("%s.snp" % base_fname, 'w')
index_var_out_file = open("%s.index.snp" % base_fname, 'w')
haplotype_out_file = open("%s.haplotype" % base_fname, 'w')
link_out_file = open("%s.link" % base_fname, 'w')
coord_out_file = open("%s.coord" % base_fname, 'w')
clnsig_out_file = open("%s.clnsig" % base_fname, 'w')
for c in range(len(chr_names)):
chr = chr_names[c]
chr_full_name = chr_full_names[c]
assert chr in chr_dic
chr_seq = chr_dic[chr]
chr_len = len(chr_seq)
if chr in genotype_genes:
chr_genes = genotype_genes[chr]
chr_genes = sorted(chr_genes, key=lambda x: (x[1], x[2], x[3]))
else:
chr_genes = []
chr_genotype_vars = []
chr_genotype_vari = 0
if graph_index:
if chr in genotype_vars:
chr_genotype_vars = genotype_vars[chr]
chr_genotype_haplotypes = []
chr_genotype_hti = 0
if chr in genotype_haplotypes:
chr_genotype_haplotypes = genotype_haplotypes[chr]
def add_vars(left, right, chr_genotype_vari, chr_genotype_hti,
haplotype_num):
# Output variants with clinical significance
while chr_genotype_vari < len(chr_genotype_vars):
var_left, \
var_type, \
var_data, \
var_id \
= chr_genotype_vars[chr_genotype_vari]
var_right = var_left
if var_type == "deletion":
var_right += var_data
if var_right > right:
break
if var_right >= left:
chr_genotype_vari += 1
continue
out_str = "%s\t%s\t%s\t%d\t%s" % (var_id, var_type, chr,
var_left + off, var_data)
print(out_str, file=var_out_file)
print(out_str, file=index_var_out_file)
if var_id in genotype_clnsig:
var_gene, clnsig = genotype_clnsig[var_id]
print("%s\t%s\t%s" \
% (var_id, var_gene, clnsig), file=clnsig_out_file)
chr_genotype_vari += 1
# Output haplotypes
while chr_genotype_hti < len(chr_genotype_haplotypes):
ht_left, ht_right, ht_vars = chr_genotype_haplotypes[
chr_genotype_hti]
if ht_right > right:
break
if ht_right >= left:
chr_genotype_hti += 1
continue
print("ht%d\t%s\t%d\t%d\t%s" \
% (haplotype_num,
chr,
ht_left + off,
ht_right + off,
','.join(ht_vars)),
file=haplotype_out_file)
chr_genotype_hti += 1
haplotype_num += 1
return chr_genotype_vari, chr_genotype_hti, haplotype_num
out_chr_seq = ""
off = 0
prev_right = 0
for gene in chr_genes:
left, right, length, name, family, exon_str, strand = gene
if not graph_index:
# Output gene (genotype_genome.gene)
print("%s\t%s\t%s\t%d\t%d\t%s\t%s" \
% (family.upper(),
name,
chr,
left,
right,
exon_str,
strand),
file=locus_out_file)
continue
chr_genotype_vari, \
chr_genotype_hti, \
haplotype_num \
= add_vars(left,
right,
chr_genotype_vari,
chr_genotype_hti,
haplotype_num)
# Read gene family sequences and information
allele_seqs = typing_common.read_allele_seq("%s_backbone.fa" %
family)
allele_vars = typing_common.read_variants("%s.snp" % family)
allele_index_vars = typing_common.read_variants("%s.index.snp" %
family)
allele_haplotypes = typing_common.read_haplotypes("%s.haplotype" %
family)
links = typing_common.read_links("%s.link" % family, True)
if name not in allele_seqs:
continue
if name not in allele_vars or name not in allele_index_vars:
vars = []
index_vars = []
else:
vars = allele_vars[name]
index_vars = allele_index_vars[name]
allele_seq = allele_seqs[name]
index_var_ids = set()
for _, _, _, var_id in index_vars:
index_var_ids.add(var_id)
if name not in allele_haplotypes:
haplotypes = []
else:
haplotypes = allele_haplotypes[name]
assert length == len(allele_seq)
assert left < chr_len and right < chr_len
# Skipping overlapping genes
if left < prev_right:
print("Warning: skipping %s ..." % (name), file=sys.stderr)
continue
varID2htID = {}
assert left < right
prev_length = right - left + 1
assert prev_length <= length
if prev_right < left:
out_chr_seq += chr_seq[prev_right:left]
# Output gene (genotype_genome.locus)
print("%s\t%s\t%s\t%d\t%d\t%s\t%s" \
% (family.upper(),
name,
chr,
len(out_chr_seq),
len(out_chr_seq) + length - 1,
exon_str,
strand),
file=locus_out_file)
# Output coord (genotype_genome.coord)
print("%s\t%d\t%d\t%d" \
% (chr,
len(out_chr_seq),
left,
right - left + 1),
file=coord_out_file)
out_chr_seq += allele_seq
# Output variants (genotype_genome.snp and genotype_genome.index.snp)
for var in vars:
var_left, var_type, var_data, var_id = var
new_var_id = "hv%d" % var_num
varID2htID[var_id] = new_var_id
new_var_left = var_left + left + off
assert var_type in ["single", "deletion", "insertion"]
assert new_var_left < len(out_chr_seq)
if var_type == "single":
assert out_chr_seq[new_var_left] != var_data
elif var_type == "deletion":
assert new_var_left + var_data <= len(out_chr_seq)
else:
assert var_type == "insertion"
out_str = "%s\t%s\t%s\t%d\t%s" \
% (new_var_id, var_type, chr, new_var_left, var_data)
print(out_str, file=var_out_file)
if var_id in index_var_ids:
print(out_str, file=index_var_out_file)
var_num += 1
# Output haplotypes (genotype_genome.haplotype)
for haplotype in haplotypes:
ht_left, ht_right, ht_vars = haplotype
new_ht_left = ht_left + left + off
assert new_ht_left < len(out_chr_seq)
new_ht_right = ht_right + left + off
assert new_ht_left <= new_ht_right
assert new_ht_right <= len(out_chr_seq)
new_ht_vars = []
for var_id in ht_vars:
assert var_id in varID2htID
new_ht_vars.append(varID2htID[var_id])
print("ht%d\t%s\t%d\t%d\t%s" \
% (haplotype_num,
chr,
new_ht_left,
new_ht_right,
','.join(new_ht_vars)),
file=haplotype_out_file)
haplotype_num += 1
# Output link information between alleles and variants (genotype_genome.link)
for link in links:
var_id, allele_names = link
if var_id not in varID2htID:
continue
new_var_id = varID2htID[var_id]
print("%s\t%s" % (new_var_id, " ".join(allele_names)),
file=link_out_file)
off += (length - prev_length)
prev_right = right + 1
if not graph_index:
continue
# Write the rest of the Vars
chr_genotype_vari, \
chr_genotype_hti, \
haplotype_num \
= add_vars(sys.maxsize,
sys.maxsize,
chr_genotype_vari,
chr_genotype_hti,
haplotype_num)
print("%s\t%d\t%d\t%d" \
% (chr,
len(out_chr_seq),
prev_right,
len(chr_seq) - prev_right),
file=coord_out_file)
out_chr_seq += chr_seq[prev_right:]
assert len(out_chr_seq) == len(chr_seq) + off
# Output chromosome sequence
print(">%s" % (chr_full_name), file=genome_out_file)
line_width = 60
for s in range(0, len(out_chr_seq), line_width):
print(out_chr_seq[s:s + line_width], file=genome_out_file)
genome_out_file.close()
locus_out_file.close()
var_out_file.close()
index_var_out_file.close()
haplotype_out_file.close()
link_out_file.close()
coord_out_file.close()
clnsig_out_file.close()
allele_out_file = open("%s.allele" % base_fname, 'w')
if graph_index:
for database in database_list:
for line in open("%s.allele" % database):
allele_name = line.strip()
print("%s\t%s" % (database.upper(), allele_name),
file=allele_out_file)
allele_out_file.close()
partial_out_file = open("%s.partial" % base_fname, 'w')
if graph_index:
for database in database_list:
for line in open("%s.partial" % database):
allele_name = line.strip()
print("%s\t%s" % (database.upper(), allele_name),
file=partial_out_file)
partial_out_file.close()
if not graph_index:
shutil.copyfile("genome.fa", "%s.fa" % base_fname)
# Index genotype_genome.fa
index_cmd = ["samtools", "faidx", "%s.fa" % base_fname]
subprocess.call(index_cmd)
# Build indexes based on the above information
if graph_index:
assert aligner == "hisat2"
build_cmd = [
"hisat2-build", "-p",
str(threads), "--snp",
"%s.index.snp" % base_fname, "--haplotype",
"%s.haplotype" % base_fname,
"%s.fa" % base_fname,
"%s" % base_fname
]
else:
assert aligner in ["hisat2", "bowtie2"]
build_cmd = [
"%s-build" % aligner, "-p" if aligner == "hisat2" else "--threads",
str(threads),
"%s.fa" % base_fname,
"%s" % base_fname
]
if verbose:
print("\tRunning:", ' '.join(build_cmd), file=sys.stderr)
subprocess.call(build_cmd,
stdout=open("/dev/null", 'w'),
stderr=open("/dev/null", 'w'))
if aligner == "hisat2":
index_fnames = ["%s.%d.ht2" % (base_fname, i + 1) for i in range(8)]
else:
index_fnames = ["%s.%d.bt2" % (base_fname, i + 1) for i in range(4)]
index_fnames += [
"%s.rev.%d.bt2" % (base_fname, i + 1) for i in range(2)
]
if not typing_common.check_files(index_fnames):
print("Error: indexing failed! "\
"Perhaps, you may have forgotten to build %s executables?" \
% aligner,
file=sys.stderr)
sys.exit(1)
def genotyping(read_dir, reference_type, region_list, num_editdist, nthreads,
max_sample, assembly, out_dir, verbose, platinum_check):
for database_name in region_list:
# Extract variants, backbone sequence, and other sequeces
typing_common.extract_database_if_not_exists(database_name,
[]) # locus_list
# Build HISAT2's graph index
typing_common.build_index_if_not_exists(
database_name,
"hisat2",
"graph",
1, # threads
verbose)
if not os.path.exists(read_dir):
print("Error: %s does not exist." % read_dir, file=sys.stderr)
sys.exit(1)
if out_dir != "" and not os.path.exists(out_dir):
os.mkdir(out_dir)
# fastq files
fq_fnames = glob.glob("%s/*.extracted.1.fq.gz" % read_dir)
genotype_results = []
lock = threading.Lock()
threads = []
for t in range(nthreads):
thread = myThread(lock, fq_fnames, reference_type, region_list,
num_editdist, max_sample, assembly, out_dir,
genotype_results, verbose)
thread.start()
threads.append(thread)
for thread in threads:
thread.join()
if platinum_check:
genotype_dic = {}
for genome, allele, abundance in genotype_results:
region, _ = allele.split('*')
if region not in genotype_dic:
genotype_dic[region] = {}
if genome not in genotype_dic[region]:
genotype_dic[region][genome] = []
if len(genotype_dic[region][genome]) >= 2:
continue
# DK - debugging purposes
# if abundance < 0.15 * 100:
# continue
genotype_dic[region][genome].append([allele, abundance])
for region, region_genotype in genotype_dic.items():
print(region, file=sys.stderr)
included = 0
total = 0
for genome, genome_alleles in region_genotype.items():
genome_alleles = set([allele for allele, _ in genome_alleles])
if "father" in CEPH_pedigree[genome]:
assert "mother" in CEPH_pedigree[genome]
parents = [
CEPH_pedigree[genome]["father"],
CEPH_pedigree[genome]["mother"]
]
else:
parents = []
parent_allele_sets = []
assert len(parents) in [0, 2]
if len(parents) == 2 \
and parents[0] in region_genotype \
and parents[1] in region_genotype:
for parent_allele, _ in region_genotype[parents[0]]:
for parent_allele2, _ in region_genotype[parents[1]]:
parent_allele_sets.append(
set([parent_allele, parent_allele2]))
print(("\t", genome, genome_alleles, parent_allele_sets),
file=sys.stderr)
if len(parent_allele_sets) > 0:
total += 1
if genome_alleles in parent_allele_sets:
included += 1
print("\t%d / %d" % (included, total), file=sys.stderr)
def build_genotype_genome(base_fname,
inter_gap,
intra_gap,
threads,
database_list,
use_clinvar,
use_commonvar,
verbose):
# Download HISAT2 index
HISAT2_fnames = ["grch38",
"genome.fa",
"genome.fa.fai"]
if not typing_common.check_files(HISAT2_fnames):
typing_common.download_genome_and_index()
# Load genomic sequences
chr_dic, chr_names, chr_full_names = typing_common.read_genome(open("genome.fa"))
genotype_vars, genotype_haplotypes, genotype_clnsig = {}, {}, {}
if use_clinvar:
# Extract variants from the ClinVar database
CLINVAR_fnames = ["clinvar.vcf.gz",
"clinvar.snp",
"clinvar.haplotype",
"clinvar.clnsig"]
if not typing_common.check_files(CLINVAR_fnames):
if not os.path.exists("clinvar.vcf.gz"):
os.system("wget ftp://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh38/archive/2017/clinvar_20170404.vcf.gz")
assert os.path.exists("clinvar.vcf.gz")
extract_cmd = ["hisat2_extract_snps_haplotypes_VCF.py"]
extract_cmd += ["--inter-gap", str(inter_gap),
"--intra-gap", str(intra_gap),
"--genotype-vcf", "clinvar.vcf.gz",
"genome.fa", "/dev/null", "clinvar"]
if verbose:
print >> sys.stderr, "\tRunning:", ' '.join(extract_cmd)
proc = subprocess.Popen(extract_cmd, stdout=open("/dev/null", 'w'), stderr=open("/dev/null", 'w'))
proc.communicate()
if not typing_common.check_files(CLINVAR_fnames):
print >> sys.stderr, "Error: extract variants from clinvar failed!"
sys.exit(1)
# Read variants to be genotyped
genotype_vars = typing_common.read_variants("clinvar.snp")
# Read haplotypes
genotype_haplotypes = typing_common.read_haplotypes("clinvar.haplotype")
# Read information about clinical significance
genotype_clnsig = typing_common.read_clnsig("clinvar.clnsig")
if use_commonvar:
# Extract variants from dbSNP database
commonvar_fbase = "snp144Common"
commonvar_fnames = ["%s.snp" % commonvar_fbase,
"%s.haplotype" % commonvar_fbase]
if not typing_common.check_files(commonvar_fnames):
if not os.path.exists("%s.txt.gz" % commonvar_fbase):
os.system("wget http://hgdownload.cse.ucsc.edu/goldenPath/hg38/database/%s.txt.gz" % commonvar_fbase)
assert os.path.exists("%s.txt.gz" % commonvar_fbase)
os.system("gzip -cd %s.txt.gz | awk 'BEGIN{OFS=\"\t\"} {if($2 ~ /^chr/) {$2 = substr($2, 4)}; if($2 == \"M\") {$2 = \"MT\"} print}' > %s.txt" % (commonvar_fbase, commonvar_fbase))
extract_cmd = ["hisat2_extract_snps_haplotypes_UCSC.py",
"--inter-gap", str(inter_gap),
"--intra-gap", str(intra_gap),
"genome.fa", "%s.txt" % commonvar_fbase, commonvar_fbase]
if verbose:
print >> sys.stderr, "\tRunning:", ' '.join(extract_cmd)
proc = subprocess.Popen(extract_cmd, stdout=open("/dev/null", 'w'), stderr=open("/dev/null", 'w'))
proc.communicate()
if not typing_common.check_files(commonvar_fnames):
print >> sys.stderr, "Error: extract variants from clinvar failed!"
sys.exit(1)
# Read variants to be genotyped
genotype_vars = typing_common.read_variants("%s.snp" % commonvar_fbase)
# Read haplotypes
genotype_haplotypes = typing_common.read_haplotypes("%s.haplotype" % commonvar_fbase)
# Genes to be genotyped
genotype_genes = {}
# Read genes or genomics regions
for database_name in database_list:
# Extract HLA variants, backbone sequence, and other sequeces
typing_common.extract_database_if_not_exists(database_name,
[], # locus_list
inter_gap,
intra_gap,
True, # partial?
verbose)
locus_fname = "%s.locus" % database_name
assert os.path.exists(locus_fname)
for line in open(locus_fname):
HLA_name, chr, left, right, length, exon_str, strand = line.strip().split()
left, right = int(left), int(right)
length = int(length)
if chr not in chr_names:
continue
if chr not in genotype_genes:
genotype_genes[chr] = []
genotype_genes[chr].append([left, right, length, HLA_name, database_name, exon_str, strand])
# Write genotype genome
var_num, haplotype_num = 0, 0
genome_out_file = open("%s.fa" % base_fname, 'w')
locus_out_file = open("%s.locus" % base_fname, 'w')
var_out_file = open("%s.snp" % base_fname, 'w')
index_var_out_file = open("%s.index.snp" % base_fname, 'w')
haplotype_out_file = open("%s.haplotype" % base_fname, 'w')
link_out_file = open("%s.link" % base_fname, 'w')
coord_out_file = open("%s.coord" % base_fname, 'w')
clnsig_out_file = open("%s.clnsig" % base_fname, 'w')
for c in range(len(chr_names)):
chr = chr_names[c]
chr_full_name = chr_full_names[c]
assert chr in chr_dic
chr_seq = chr_dic[chr]
chr_len = len(chr_seq)
if chr in genotype_genes:
chr_genes = genotype_genes[chr]
def gene_cmp(a, b):
a_left, a_right, a_length = a[:3]
b_left, b_right, b_length = b[:3]
if a_left != b_left:
return a_left - b_left
if a_right != b_right:
return a_right - b_right
return a_lenght - b_length
chr_genes = sorted(chr_genes, cmp=gene_cmp)
else:
chr_genes = []
chr_genotype_vars, chr_genotype_vari = [], 0
if chr in genotype_vars:
chr_genotype_vars = genotype_vars[chr]
chr_genotype_haplotypes, chr_genotype_hti = [], 0
if chr in genotype_haplotypes:
chr_genotype_haplotypes = genotype_haplotypes[chr]
def add_vars(left, right, chr_genotype_vari, chr_genotype_hti, haplotype_num):
# Output variants with clinical significance
while chr_genotype_vari < len(chr_genotype_vars):
var_left, var_type, var_data, var_id = chr_genotype_vars[chr_genotype_vari]
var_right = var_left
if var_type == "deletion":
var_right += var_data
if var_right > right:
break
if var_right >= left:
chr_genotype_vari += 1
continue
out_str = "%s\t%s\t%s\t%d\t%s" % (var_id, var_type, chr, var_left + off, var_data)
print >> var_out_file, out_str
print >> index_var_out_file, out_str
if var_id in genotype_clnsig:
var_gene, clnsig = genotype_clnsig[var_id]
print >> clnsig_out_file, "%s\t%s\t%s" % \
(var_id, var_gene, clnsig)
chr_genotype_vari += 1
# Output haplotypes
while chr_genotype_hti < len(chr_genotype_haplotypes):
ht_left, ht_right, ht_vars = chr_genotype_haplotypes[chr_genotype_hti]
if ht_right > right:
break
if ht_right >= left:
chr_genotype_hti += 1
continue
print >> haplotype_out_file, "ht%d\t%s\t%d\t%d\t%s" % \
(haplotype_num, chr, ht_left + off, ht_right + off, ','.join(ht_vars))
chr_genotype_hti += 1
haplotype_num += 1
return chr_genotype_vari, chr_genotype_hti, haplotype_num
out_chr_seq = ""
off = 0
prev_right = 0
for gene in chr_genes:
left, right, length, name, family, exon_str, strand = gene
chr_genotype_vari, chr_genotype_hti, haplotype_num = add_vars(left, right, chr_genotype_vari, chr_genotype_hti, haplotype_num)
# Read HLA backbone sequences
allele_seqs = typing_common.read_allele_sequences("%s_backbone.fa" % family)
# Read HLA variants
allele_vars = typing_common.read_variants("%s.snp" % family)
allele_index_vars = typing_common.read_variants("%s.index.snp" % family)
# Read HLA haplotypes
allele_haplotypes = typing_common.read_haplotypes("%s.haplotype" % family)
# Read HLA link information between haplotypes and variants
links = typing_common.read_links("%s.link" % family)
if name not in allele_seqs or \
name not in allele_vars or \
name not in allele_haplotypes:
continue
allele_seq = allele_seqs[name]
vars, index_vars = allele_vars[name], allele_index_vars[name]
index_var_ids = set()
for _, _, _, var_id in index_vars:
index_var_ids.add(var_id)
haplotypes = allele_haplotypes[name]
assert length == len(allele_seq)
assert left < chr_len and right < chr_len
# Skipping overlapping genes
if left < prev_right:
print >> sys.stderr, "Warning: skipping %s ..." % (name)
continue
varID2htID = {}
assert left < right
prev_length = right - left + 1
assert prev_length <= length
if prev_right < left:
out_chr_seq += chr_seq[prev_right:left]
# Output gene (genotype_genome.gene)
print >> locus_out_file, "%s\t%s\t%s\t%d\t%d\t%s\t%s" % \
(family.upper(), name, chr, len(out_chr_seq), len(out_chr_seq) + length - 1, exon_str, strand)
# Output coord (genotype_genome.coord)
print >> coord_out_file, "%s\t%d\t%d\t%d" % \
(chr, len(out_chr_seq), left, right - left + 1)
out_chr_seq += allele_seq
# Output variants (genotype_genome.snp and genotype_genome.index.snp)
for var in vars:
var_left, var_type, var_data, var_id = var
new_var_id = "hv%d" % var_num
varID2htID[var_id] = new_var_id
new_var_left = var_left + left + off
assert var_type in ["single", "deletion", "insertion"]
assert new_var_left < len(out_chr_seq)
if var_type == "single":
assert out_chr_seq[new_var_left] != var_data
elif var_type == "deletion":
assert new_var_left + var_data <= len(out_chr_seq)
else:
assert var_type == "insertion"
out_str = "%s\t%s\t%s\t%d\t%s" % (new_var_id, var_type, chr, new_var_left, var_data)
print >> var_out_file, out_str
if var_id in index_var_ids:
print >> index_var_out_file, out_str
var_num += 1
# Output haplotypes (genotype_genome.haplotype)
for haplotype in haplotypes:
ht_left, ht_right, ht_vars = haplotype
new_ht_left = ht_left + left + off
assert new_ht_left < len(out_chr_seq)
new_ht_right = ht_right + left + off
assert new_ht_left <= new_ht_right
assert new_ht_right <= len(out_chr_seq)
new_ht_vars = []
for var_id in ht_vars:
assert var_id in varID2htID
new_ht_vars.append(varID2htID[var_id])
print >> haplotype_out_file, "ht%d\t%s\t%d\t%d\t%s" % \
(haplotype_num, chr, new_ht_left, new_ht_right, ','.join(new_ht_vars))
haplotype_num += 1
# Output link information between alleles and variants (genotype_genome.link)
for link in links:
var_id, allele_names = link
if var_id not in varID2htID:
continue
new_var_id = varID2htID[var_id]
print >> link_out_file, "%s\t%s" % (new_var_id, allele_names)
off += (length - prev_length)
prev_right = right + 1
# Write the rest of the Vars
chr_genotype_vari, chr_genotype_hti, haplotype_num = add_vars(sys.maxint, sys.maxint, chr_genotype_vari, chr_genotype_hti, haplotype_num)
print >> coord_out_file, "%s\t%d\t%d\t%d" % \
(chr, len(out_chr_seq), prev_right, len(chr_seq) - prev_right)
out_chr_seq += chr_seq[prev_right:]
assert len(out_chr_seq) == len(chr_seq) + off
# Output chromosome sequence
print >> genome_out_file, ">%s" % (chr_full_name)
line_width = 60
for s in range(0, len(out_chr_seq), line_width):
print >> genome_out_file, out_chr_seq[s:s+line_width]
genome_out_file.close()
locus_out_file.close()
var_out_file.close()
index_var_out_file.close()
haplotype_out_file.close()
link_out_file.close()
coord_out_file.close()
clnsig_out_file.close()
partial_out_file = open("%s.partial" % base_fname, 'w')
for database in database_list:
for line in open("%s.partial" % database):
allele_name = line.strip()
print >> partial_out_file, "%s\t%s" % (database.upper(), allele_name)
partial_out_file.close()
# Index genotype_genome.fa
index_cmd = ["samtools", "faidx", "%s.fa" % base_fname]
subprocess.call(index_cmd)
# Build HISAT-genotype graph indexes based on the above information
hisat2_index_fnames = ["%s.%d.ht2" % (base_fname, i+1) for i in range(8)]
build_cmd = ["hisat2-build",
"-p", str(threads),
"--snp", "%s.index.snp" % base_fname,
"--haplotype", "%s.haplotype" % base_fname,
"%s.fa" % base_fname,
"%s" % base_fname]
if verbose:
print >> sys.stderr, "\tRunning:", ' '.join(build_cmd)
subprocess.call(build_cmd, stdout=open("/dev/null", 'w'), stderr=open("/dev/null", 'w'))
if not typing_common.check_files(hisat2_index_fnames):
print >> sys.stderr, "Error: indexing failed! Perhaps, you may have forgotten to build hisat2 executables?"
sys.exit(1)