def ace2fasta(in_file, out_file):
ace_gen = Ace.parse(open(in_file, 'r'))
with open(out_file, "w") as output_file:
while 1:
try:
contig = ace_gen.next()
except:
print "All contigs treated"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
# Now we have started our alignment we can add sequences to it
# Add concensus sequence to alignment
align.add_sequence(contig.name, contig.sequence)
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
output_file.write(align.format("fasta"))
def __init__(self,ace_file):
self.ace_file = ace_file
self.records = Ace.read(open(ace_file, 'r'))
assert len(self.records.contigs)==1
self.contig = self.records.contigs[0]
self.consensus = self.contig.sequence
self.consensus_name = self.contig.name
self.number_sequences = len(self.contig.reads)
self.reference = ""
self.reference_name = ""
def gene_expression_2matrix(in_ace, out_file, tags, min_seq):
"""Count sequences with each tags in all contigs.
"""
print
print "USING MATRIX OUTPUT FORMAT"
print
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
output_file.write("gene_name\tgene_length")
for tag in tags:
output_file.write("\t" + tag)
output_file.write("\tXX_noTag")
output_file.write("\n")
while 1:
try:
contig = ace_gen.next()
except:
print "***All contigs treated***"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
align.add_sequence(contig.name, contig.sequence)
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta_2list(align.format("fasta"))
if len(sequences) < min_seq:
continue
contig_name = re.findall("(Contig_[0-9]+)", sequences[0][0])[0]
contig_seq = sequences[0][1].replace("*", "")
contig_length = str(len(contig_seq))
output_file.write(contig_name + "\t" + contig_length)
print "Treating", contig_name
d = defaultdict(int)
for tag in tags:
d[tag] = 0
d["XX_noTag"] = 0
fasta_counter = 0
for fasta in sequences:
fasta_counter += 1
found_tag = 0
for tag in tags:
if fasta[0].find(tag) > -1:
d[tag] += 1
found_tag = 1
if found_tag == 0 and fasta[0].find("Consensus") < 0:
d["XX_noTag"] += 1
for tag in sorted(d):
output_file.write("\t" + str(d[tag]))
output_file.write("\n")
def report(args):
"""
%prog report [--options] ace_file > report
Prepare a report of read location, consensus location or quality segment per contig
"""
from jcvi.utils.table import tabulate
p = OptionParser(report.__doc__)
types = {"read": ["padded_start", "padded_end", "orient"],
"consensus": ["padded_consensus_start", "padded_consensus_end"],
"quality" : ["qual_clipping_start", "qual_clipping_end", "align_clipping_start", "align_clipping_end"]
}
valid_types = tuple(types.keys())
p.add_option("--type", default="read", choices=valid_types,
help="choose report type [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
acefile, = args
ace = Ace.read(must_open(acefile))
logging.debug('Loaded ace file {0}'.format(acefile))
for c in ace.contigs:
print c.name
table = dict()
if opts.type == "read":
ps, pe = [], []
ps = [read.padded_start for read in c.af]
for i in xrange(1, len(ps)):
pe.append(ps[i] - ps[i-1])
pe.append(c.nbases)
map = dict(zip(ps, pe))
for i, read in enumerate(c.af):
values = [str(x) for x in (read.padded_start, map[read.padded_start], read.coru)]
for i, label in enumerate(types[opts.type]):
table[(str(read.name), label)] = values[i]
elif opts.type == "consensus":
for read in c.bs:
values = [str(x) for x in (read.padded_start, read.padded_end)]
for i, label in enumerate(types[opts.type]):
table[(str(read.name), label)] = values[i]
elif opts.type == "quality":
for read in c.reads:
(r1, r2) = (read.rd, read.qa)
values = [str(x) for x in (r2.qual_clipping_start, r2.qual_clipping_end, r2.align_clipping_start, r2.align_clipping_end)]
for i, label in enumerate(types[opts.type]):
table[(str(r1.name), label)] = values[i]
print tabulate(table), "\n"
def main():
base_name = 'FX5ZTWB02D1DFX'
#contigs = Ace.parse(open('/Users/bcf/Tmp/tmp2.fa.cap.ace'))
c = Ace.read(open('/Users/bcf/Tmp/tmp2.fa.cap.ace'))
'''for c in contigs:
for r in c.reads:
if r.rd.name == base_name:
contig = c
break
else:
pass'''
write(c, '/Users/bcf/Tmp/tmp2_rewrite.fa.cap.ace')
pdb.set_trace()
def parse_ace(ace_file): ace_gen = Ace.parse(open(ace_file, 'r')) contig = ace_gen.next() align = Alignment(Gapped(IUPAC.ambiguous_dna, "-")) align.add_sequence(contig.name, contig.sequence) for readn in range(len(contig.reads)): clipst = contig.reads[readn].qa.qual_clipping_start clipe = contig.reads[readn].qa.qual_clipping_end start = contig.af[readn].padded_start seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe) seq = pad_read(seq, start, len(contig.sequence)) align.add_sequence(contig.reads[readn].rd.name + "_" + contig.af[readn].coru, seq) return contig, align
def extract(args):
"""
%prog extract [--options] ace_file
Extract contigs from ace file and if necessary reformat header with
a pipe(|) separated list of constituent reads.
"""
p = OptionParser(extract.__doc__)
p.add_option("--format", default=False, action="store_true",
help="enable flag to reformat header into a symbol separated list of constituent reads "+ \
"[default: %default]")
p.add_option("--sep", default="|",
help="choose a separator used to list the reads in the FASTA header [default: '%default']")
p.add_option("--singlets", default=False, action="store_true",
help="ask the program to look in the singlets file (should be in the same folder) for " +\
"unused reads and put them in the resultant fasta file [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
acefile, = args
ace = Ace.read(must_open(acefile))
logging.debug('Loaded ace file {0}'.format(acefile))
fastafile = acefile.rsplit(".", 1)[0] + ".fasta"
fw = open(fastafile, "w")
for c in ace.contigs:
id = c.name
if opts.format:
id = opts.sep.join([read.name for read in c.af])
seqrec = SeqRecord(Seq(c.sequence), id=id, description="")
SeqIO.write([seqrec], fw, "fasta")
if opts.singlets:
singletsfile = acefile.rsplit(".", 1)[0] + ".singlets"
if os.path.getsize(singletsfile) > 0:
fp = SeqIO.parse(must_open(singletsfile), "fasta")
for rec in fp:
SeqIO.write(rec, fw, "fasta")
fw.close()
logging.debug('Wrote contigs to fasta file {0}'.format(fastafile))
def extract(args):
"""
%prog extract [--options] ace_file
Extract contigs from ace file and if necessary reformat header with
a pipe(|) separated list of constituent reads.
"""
p = OptionParser(extract.__doc__)
p.add_option("--format", default=False, action="store_true",
help="enable flag to reformat header into a symbol separated list of constituent reads "+ \
"[default: %default]")
p.add_option("--sep", default="|",
help="choose a separator used to list the reads in the FASTA header [default: '%default']")
p.add_option("--singlets", default=False, action="store_true",
help="ask the program to look in the singlets file (should be in the same folder) for " +\
"unused reads and put them in the resultant fasta file [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
acefile, = args
ace = Ace.read(must_open(acefile))
logging.debug('Loaded ace file {0}'.format(acefile))
fastafile = acefile.rsplit(".", 1)[0] + ".fasta"
fw = open(fastafile, "w")
for c in ace.contigs:
id = c.name
if opts.format:
id = opts.sep.join([read.name for read in c.af])
seqrec = SeqRecord(Seq(c.sequence), id=id, description="")
SeqIO.write([seqrec], fw, "fasta")
if opts.singlets:
singletsfile = acefile.rsplit(".", 1)[0] + ".singlets"
if os.path.getsize(singletsfile) > 0:
fp = SeqIO.parse(must_open(singletsfile), "fasta")
for rec in fp:
SeqIO.write(rec, fw, "fasta")
fw.close()
logging.debug('Wrote contigs to fasta file {0}'.format(fastafile))
def AceIterator(handle) :
"""Returns SeqRecord objects from an ACE file.
This uses the Bio.Sequencing.Ace module to do the hard work. Note that
by iterating over the file in a single pass, we are forced to ignore any
WA, CT, RT or WR footer tags."""
for ace_contig in Ace.parse(handle) :
#Convert the ACE contig record into a SeqRecord...
consensus_seq_str = ace_contig.sequence
#Assume its DNA unless there is a U in it,
if "U" in consensus_seq_str :
if "T" in consensus_seq_str :
#Very odd! Error?
alpha = generic_ncleotide
else :
alpha = generic_rna
else :
alpha = generic_dna
if "*" in consensus_seq_str :
#For consistency with most other file formats, map
#any * gaps into 0 gaps.
assert "-" not in consensus_seq_str
consensus_seq = Seq(consensus_seq_str.replace("*","-"),
Gapped(alpha, gap_char="-"))
else :
consensus_seq = Seq(consensus_seq_str, alpha)
#TODO - Consensus base quality (BQ lines). Note that any gaps
#(* character) in the consensus does not get a quality entry.
#This really needs Biopython support for per-letter-annotation.
#TODO? - Base segments (BS lines) which indicates which read
#phrap has chosen to be the consensus at a particular position.
#Perhaps as SeqFeature objects?
#TODO - Supporting reads (RD lines, plus perhaps QA and DS lines)
#Perhaps as SeqFeature objects?
seq_record = SeqRecord(consensus_seq,
id = ace_contig.name,
name = ace_contig.name)
yield seq_record
def AceIterator(handle):
"""Returns SeqRecord objects from an ACE file.
This uses the Bio.Sequencing.Ace module to do the hard work. Note that
by iterating over the file in a single pass, we are forced to ignore any
WA, CT, RT or WR footer tags."""
for ace_contig in Ace.parse(handle):
#Convert the ACE contig record into a SeqRecord...
consensus_seq_str = ace_contig.sequence
#Assume its DNA unless there is a U in it,
if "U" in consensus_seq_str:
if "T" in consensus_seq_str:
#Very odd! Error?
alpha = generic_ncleotide
else:
alpha = generic_rna
else:
alpha = generic_dna
if "*" in consensus_seq_str:
#For consistency with most other file formats, map
#any * gaps into 0 gaps.
assert "-" not in consensus_seq_str
consensus_seq = Seq(consensus_seq_str.replace("*", "-"),
Gapped(alpha, gap_char="-"))
else:
consensus_seq = Seq(consensus_seq_str, alpha)
#TODO - Consensus base quality (BQ lines). Note that any gaps
#(* character) in the consensus does not get a quality entry.
#This really needs Biopython support for per-letter-annotation.
#TODO? - Base segments (BS lines) which indicates which read
#phrap has chosen to be the consensus at a particular position.
#Perhaps as SeqFeature objects?
#TODO - Supporting reads (RD lines, plus perhaps QA and DS lines)
#Perhaps as SeqFeature objects?
seq_record = SeqRecord(consensus_seq,
id=ace_contig.name,
name=ace_contig.name)
yield seq_record
def parse_singletons_fasta_in_ace(contig_ace_dir, singleton_seq_dir):
# get contig info
os.chdir(contig_ace_dir)
for ace_file in sorted(glob.glob("*.ace")):
ace_record = Ace.read(open(ace_file))
contigs = ace_record.contigs
for contig in contigs:
if contig.nreads == 1:
singleton_name = contig.reads[0].rd.name
singleton_seq = Seq(contig.reads[0].rd.sequence)
singleton_record = SeqRecord(seq=singleton_seq,
id="",
name="",
description=singleton_name)
singleton_file = singleton_seq_dir + "/" + singleton_name + ".fsa"
singleton_fd = open(singleton_file, "w")
SeqIO.write([singleton_record], singleton_fd, "fasta")
singleton_fd.close()
os.system("sed -i \"s/> />/g\" " + singleton_file)
def ace2fasta(in_file, out_file):
ace_gen = Ace.parse(open(in_file, 'r'))
with open(out_file, "w") as output_file:
while 1:
try:
contig = ace_gen.next()
except:
print "All contigs treated"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
# Now we have started our alignment we can add sequences to it
# Add concensus sequence to alignment
align.add_sequence(contig.name, contig.sequence.replace("*", ""))
"""for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)"""
output_file.write(align.format("fasta"))
from Bio.Sequencing import Ace
fn = '../../samples/contig1.ace'
acefilerecord = Ace.read(open(fn))
# For each contig:
for ctg in acefilerecord.contigs:
print('==========================================')
print('Contig name: %s'%ctg.name)
print('Bases: %s'%ctg.nbases)
print('Reads: %s'%ctg.nreads)
print('Segments: %s'%ctg.nsegments)
print('Sequence: %s'%ctg.sequence)
print('Quality: %s'%ctg.quality)
# For each read in contig:
for read in ctg.reads:
print('Read name: %s'%read.rd.name)
print('Align start: %s'%read.qa.align_clipping_start)
print('Align end: %s'%read.qa.align_clipping_end)
print('Qual start: %s'%read.qa.qual_clipping_start)
print('Qual end: %s'%read.qa.qual_clipping_end)
print('Read sequence: %s'%read.rd.sequence)
print('==========================================')
def pairwise(in_ace, out_file):
"""Calculate pairwise differentiation indexes.
"""
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
while 1:
try:
contig = ace_gen.next()
except:
print "***All contigs treated***"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
align.add_sequence(contig.name, contig.sequence)
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta(align.format("fasta"))
contig_name = re.findall("(Contig_[0-9]+)", sequences[0][0])[0]
print "Treating", contig_name
window_len = 8 # PARAMETER
max_diff = 3 # PARAMETER
len_contig = len(sequences[0][1])
number_indexes = 0
total_indexes = 0
for seq in sequences[1:]:
try:
start = len(re.findall("^-+", seq[1])[0])
except:
start = 0
len_seq = 0
min_len_seq = 100 # PARAMETER
count = 0
for window in range(start, len_contig, window_len):
nuc_contig = sequences[0][1][window:window + window_len]
nuc_seq = seq[1][window:window + window_len]
if "-" in nuc_seq:
len_seq += len(nuc_seq.replace("-", ""))
else:
diff = count_diff(nuc_contig, nuc_seq, max_diff)
if diff[1] == False:
count += diff[0]
len_seq += window_len
len_seq -= seq.count("*")
if len_seq >= min_len_seq:
index = float(count) / len_seq
if count > 0:
number_indexes += 1
total_indexes += index
else:
index = "NA"
#output_file.write(contig_name + "\t" + str(index) + "\n")
try:
mean_index = float(total_indexes) / number_indexes
except:
mean_index = "NA"
output_file.write(contig_name + "\t" + str(mean_index) + "\n")
def pairwise(in_ace, out_file):
"""Calculate pairwise differentiation indexes.
"""
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
while 1:
try:
contig = ace_gen.next()
except:
print "***All contigs treated***"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
align.add_sequence(contig.name, contig.sequence)
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta(align.format("fasta"))
contig_name = re.findall("(Contig_[0-9]+)", sequences[0][0])[0]
print "Treating", contig_name
window_len = 8 # PARAMETER
max_diff = 3 # PARAMETER
len_contig = len(sequences[0][1])
number_indexes = 0
total_indexes = 0
for seq in sequences[1:]:
try:
start = len(re.findall("^-+", seq[1])[0])
except:
start = 0
len_seq = 0
min_len_seq = 100 # PARAMETER
count = 0
for window in range(start, len_contig, window_len):
nuc_contig = sequences[0][1][window:window + window_len]
nuc_seq = seq[1][window:window + window_len]
if "-" in nuc_seq:
len_seq += len(nuc_seq.replace("-", ""))
else:
diff = count_diff(nuc_contig, nuc_seq, max_diff)
if diff[1] == False:
count += diff[0]
len_seq += window_len
len_seq -= seq.count("*")
if len_seq >= min_len_seq:
index = float(count) / len_seq
if count > 0:
number_indexes +=1
total_indexes += index
else:
index = "NA"
#output_file.write(contig_name + "\t" + str(index) + "\n")
try:
mean_index = float(total_indexes) / number_indexes
except:
mean_index = "NA"
output_file.write(contig_name + "\t" + str(mean_index) + "\n")
def get_haplotypes(in_ace, out_file, out_bamova, win_len, step,
coverage, stars, ngroups, nhaplo):
"""Get haplotypes from contigs in an ace file
"""
marker_number = 0
min_freq = 0.05
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
with open(out_bamova, "w") as bamova_file:
output_file.write("Contig_nb\tWindow\tHaplotype\n")
contig_counter = 0
ntreated = 0
for contig in ace_gen:
pass_haplo = False
contig_counter += 1
align = Alignment(Gapped(IUPAC.ambiguous_dna, "X"))
align.add_sequence(contig.name, contig.sequence)
if len(contig.reads) -1 < coverage:
continue
ntreated += 1
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
clipst2 = contig.reads[readn].qa.align_clipping_start
clipe2 = contig.reads[readn].qa.align_clipping_end
if clipst2 > clipst:
clipst = clipst2
if clipe2 < clipe2:
clipe = clipe2
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta(align.format("fasta"))
sequences = [[s[0].replace(">", ""), s[1]] for s in sequences]
contig_name = sequences[0][0]
concensus = sequences[0][1]
error_positions = multi_find("*", concensus)[::-1]
for p in error_positions:
sequences = [[s[0], s[1][0:p] + s[1][p+1:]] for s in sequences]
concensus = sequences[0][1]
sequences = [[s[0], correct_sequence(concensus, s[1])]
for s in sequences[1:]]
sequences, snp_pos = snp_positions(sequences)
haplotypes = best_snps(sequences, snp_pos, coverage)
if haplotypes != "Empty":
bamova = []
variants = list(sorted(list(set([h[-1] for h in haplotypes[-1]]))))
groups = list(sorted(set([h[0][:3] for h in haplotypes[-1]])))
if len(groups) >= ngroups:
pass_haplo = True
for g in groups:
if len([h[0] for h in haplotypes[-1] if h[0].startswith(g)]) < nhaplo:
pass_haplo = False
if pass_haplo:
print contig.name
bamova_file.write("Marker" + str(marker_number) + "\n")
group_number = 0
for g in groups:
bamova_file.write("Population\t" + str(group_number))
group_number += 1
for v in variants:
bamova_file.write("\t" + str(len([h for h in haplotypes[-1]
if h[-1] == v and h[0].startswith(g)])))
bamova_file.write("\n")
with open ("fasta_output/" + contig.name + ".fasta", "w") as f:
output_file.write(contig.name + "\n")
for h in haplotypes[-1]:
f.write(">" + h[0] + str(marker_number) + "\n" + h[2] + "\n")
h[1] = [x - h[1][0] + 1 for x in h[1]]
output_file.write("Marker" + str(marker_number) + "\t" +
"\t".join([str(x) for x in h]) + "\t" +
":".join(variants) + "\n")
marker_number += 1
output_file.flush()
bamova_file.flush()
cutoff = 100000
if contig_counter > cutoff:
break
print "\n", str(ntreated), "contigs out of", str(contig_counter), "were treated"
def report(args):
"""
%prog report [--options] ace_file > report
Prepare a report of read location, consensus location or quality segment per contig
"""
from jcvi.utils.table import tabulate
p = OptionParser(report.__doc__)
types = {
"read": ["padded_start", "padded_end", "orient"],
"consensus": ["padded_consensus_start", "padded_consensus_end"],
"quality": [
"qual_clipping_start", "qual_clipping_end", "align_clipping_start",
"align_clipping_end"
]
}
valid_types = tuple(types.keys())
p.add_option("--type",
default="read",
choices=valid_types,
help="choose report type [default: %default]")
opts, args = p.parse_args(args)
if len(args) != 1:
sys.exit(not p.print_help())
acefile, = args
ace = Ace.read(must_open(acefile))
logging.debug('Loaded ace file {0}'.format(acefile))
for c in ace.contigs:
print c.name
table = dict()
if opts.type == "read":
ps, pe = [], []
ps = [read.padded_start for read in c.af]
for i in xrange(1, len(ps)):
pe.append(ps[i] - ps[i - 1])
pe.append(c.nbases)
map = dict(zip(ps, pe))
for i, read in enumerate(c.af):
values = [
str(x) for x in (read.padded_start, map[read.padded_start],
read.coru)
]
for i, label in enumerate(types[opts.type]):
table[(str(read.name), label)] = values[i]
elif opts.type == "consensus":
for read in c.bs:
values = [str(x) for x in (read.padded_start, read.padded_end)]
for i, label in enumerate(types[opts.type]):
table[(str(read.name), label)] = values[i]
elif opts.type == "quality":
for read in c.reads:
(r1, r2) = (read.rd, read.qa)
values = [
str(x)
for x in (r2.qual_clipping_start, r2.qual_clipping_end,
r2.align_clipping_start, r2.align_clipping_end)
]
for i, label in enumerate(types[opts.type]):
table[(str(r1.name), label)] = values[i]
print tabulate(table), "\n"
part_site_comp_fh.writerow(row)
cutoff = []
if (p.use_reads):
print "Working on ace file {}".format(p.read_fn)
contig_read_dict = {}
contig_read_len_dict = {}
from Bio.Sequencing import Ace
with open(p.use_reads, 'rU') as ace_fh:
for contig in Ace.parse(ace_fh):
"""rd (reads) - read with name, sequence, etc
qa (read qual) - which parts used as consensus
ds - file name of read's chromatogram file
af - loc of read within contig
bs (base segment) - which read chosen at consensus at each pos
rt (transient read tags) - generated by crossmatch and phrap
ct (consensus tag)
wa (whole assembly tag) - hosts assembly program name, version, etc
wr
reads - info about read supporting ace contig
contig - holds info about contig from ace record"""
contig_name = "{}".format(contig.name) # contig00001
if not contig_name in contig_read_dict:
def get_haplotypes(in_ace, out_file, out_bamova, win_len, step, coverage,
stars, ngroups, nhaplo):
"""Get haplotypes from contigs in an ace file
"""
marker_number = 0
min_freq = 0.05
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
with open(out_bamova, "w") as bamova_file:
output_file.write("Contig_nb\tWindow\tHaplotype\n")
contig_counter = 0
ntreated = 0
for contig in ace_gen:
pass_haplo = False
contig_counter += 1
align = Alignment(Gapped(IUPAC.ambiguous_dna, "X"))
align.add_sequence(contig.name, contig.sequence)
if len(contig.reads) - 1 < coverage:
continue
ntreated += 1
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start
clipe = contig.reads[readn].qa.qual_clipping_end
clipst2 = contig.reads[readn].qa.align_clipping_start
clipe2 = contig.reads[readn].qa.align_clipping_end
if clipst2 > clipst:
clipst = clipst2
if clipe2 < clipe2:
clipe = clipe2
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst,
clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta(align.format("fasta"))
sequences = [[s[0].replace(">", ""), s[1]] for s in sequences]
contig_name = sequences[0][0]
concensus = sequences[0][1]
error_positions = multi_find("*", concensus)[::-1]
for p in error_positions:
sequences = [[s[0], s[1][0:p] + s[1][p + 1:]]
for s in sequences]
concensus = sequences[0][1]
sequences = [[s[0], correct_sequence(concensus, s[1])]
for s in sequences[1:]]
sequences, snp_pos = snp_positions(sequences)
haplotypes = best_snps(sequences, snp_pos, coverage)
if haplotypes != "Empty":
bamova = []
variants = list(
sorted(list(set([h[-1] for h in haplotypes[-1]]))))
groups = list(
sorted(set([h[0][:3] for h in haplotypes[-1]])))
if len(groups) >= ngroups:
pass_haplo = True
for g in groups:
if len([
h[0] for h in haplotypes[-1]
if h[0].startswith(g)
]) < nhaplo:
pass_haplo = False
if pass_haplo:
print contig.name
bamova_file.write("Marker" + str(marker_number) + "\n")
group_number = 0
for g in groups:
bamova_file.write("Population\t" +
str(group_number))
group_number += 1
for v in variants:
bamova_file.write("\t" + str(
len([
h for h in haplotypes[-1]
if h[-1] == v and h[0].startswith(g)
])))
bamova_file.write("\n")
with open("fasta_output/" + contig.name + ".fasta",
"w") as f:
output_file.write(contig.name + "\n")
for h in haplotypes[-1]:
f.write(">" + h[0] + str(marker_number) +
"\n" + h[2] + "\n")
h[1] = [x - h[1][0] + 1 for x in h[1]]
output_file.write(
"Marker" + str(marker_number) + "\t" +
"\t".join([str(x) for x in h]) + "\t" +
":".join(variants) + "\n")
marker_number += 1
output_file.flush()
bamova_file.flush()
cutoff = 100000
if contig_counter > cutoff:
break
print "\n", str(ntreated), "contigs out of", str(
contig_counter), "were treated"
from Bio.Sequencing import Ace
fn = '../../samples/contig1.ace'
acefilerecord = Ace.read(open(fn))
# For each contig:
for ctg in acefilerecord.contigs:
print('==========================================')
print('Contig name: %s' % ctg.name)
print('Bases: %s' % ctg.nbases)
print('Reads: %s' % ctg.nreads)
print('Segments: %s' % ctg.nsegments)
print('Sequence: %s' % ctg.sequence)
print('Quality: %s' % ctg.quality)
# For each read in contig:
for read in ctg.reads:
print('Read name: %s' % read.rd.name)
print('Align start: %s' % read.qa.align_clipping_start)
print('Align end: %s' % read.qa.align_clipping_end)
print('Qual start: %s' % read.qa.qual_clipping_start)
print('Qual end: %s' % read.qa.qual_clipping_end)
print('Read sequence: %s' % read.rd.sequence)
print('==========================================')
def _get_gen(self):
return ace.parse(open(self.ace_filename))
def snp_count(in_ace, out_file, snp_dict, tags, win_len, max_del, stars):
"""Genotype individuals at SNPs loci.
"""
win_buffer = (win_len - 1) / 2
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
output_file.write("Contig_nb\tPos\ttag_name\tA\tC\tG\tT\tN\t*\t-\n")
while 1:
try:
contig = ace_gen.next()
except:
print "***All contigs treated***"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
align.add_sequence(contig.name, contig.sequence)
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start # GOOD
clipe = contig.reads[readn].qa.qual_clipping_end # GOOD
clipst2 = contig.reads[readn].qa.align_clipping_start # Added
clipe2 = contig.reads[readn].qa.align_clipping_end # Added
if clipst2 > clipst: # Added
clipst = clipst2 # Added
if clipe2 < clipe2: # Added
clipe = clipe2 # Added
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta(align.format("fasta"))
contig_name = re.findall("(Contig_[0-9]+)", sequences[0][0])[0]
print "Treating", contig_name
positions = []
try:
positions = snp_dict[contig_name]
except:
continue
d = {}
for pos in positions:
if stars == True:
pos_ok = correct_position(pos, sequences[0][1])
else:
pos_ok = pos
left = pos_ok - 5
if left < 0:
left = 0
right = pos_ok + 1 + 5 # takes into account the middle nucleotide
ref_window = sequences[0][1][left:right]
d.setdefault(pos, {})
d[pos].setdefault("XX_noTag", {})
for nuc in list("ACGTN*-"):
d[pos]["XX_noTag"].setdefault(nuc, 0)
for tag in tags:
d[pos].setdefault(tag, {})
for nuc in list("ACGTN*-"):
d[pos][tag].setdefault(nuc, 0)
for fasta in sequences:
window = fasta[1][left:right]
del_count = 0
if window.count("-") > win_buffer - 3:
continue # Need at least 3 nucleotides on each side
for tag in tags:
if tag in fasta[0]:
t = tag
break
else:
t = "XX_noTag"
if len(ref_window) == len(window):
for i in xrange(len(window)):
if ref_window[i].isalpha() and window[i] == "*" or \
window[i].isalpha() and ref_window[i] == "*":
del_count += 1
if del_count > max_del:
continue
p = pos
s = fasta[1] # Sequence
n = s[pos_ok - 1].upper()
d[p][t][n] += 1
for p in sorted(d):
for t in sorted(d[p]):
output_file.write(contig_name + "\t" + str(p) + "\t" +
str(t))
for n in list("ACGTN*-"):
output_file.write("\t" + str(d[p][t][n]))
output_file.write("\n")
def AceIterator(handle):
"""Returns SeqRecord objects from an ACE file.
This uses the Bio.Sequencing.Ace module to do the hard work. Note that
by iterating over the file in a single pass, we are forced to ignore any
WA, CT, RT or WR footer tags.
Ace files include the base quality for each position, which are taken
to be PHRED style scores. Just as if you had read in a FASTQ or QUAL file
using PHRED scores using Bio.SeqIO, these are stored in the SeqRecord's
letter_annotations dictionary under the "phred_quality" key.
>>> from Bio import SeqIO
>>> handle = open("Ace/consed_sample.ace", "rU")
>>> for record in SeqIO.parse(handle, "ace"):
... print record.id, record.seq[:10]+"...", len(record)
... print max(record.letter_annotations["phred_quality"])
Contig1 agccccgggc... 1475
90
However, ACE files do not include a base quality for any gaps in the
consensus sequence, and these are represented in Biopython with a quality
of zero. Using zero is perhaps misleading as there may be very strong
evidence to support the gap in the consensus. Previous versions of
Biopython therefore used None instead, but this complicated usage, and
prevented output of the gapped sequence as FASTQ format.
>>> from Bio import SeqIO
>>> handle = open("Ace/contig1.ace", "rU")
>>> for record in SeqIO.parse(handle, "ace"):
... print record.id, "..." + record.seq[85:95]+"..."
... print record.letter_annotations["phred_quality"][85:95]
... print max(record.letter_annotations["phred_quality"])
Contig1 ...AGAGG-ATGC...
[57, 57, 54, 57, 57, 0, 57, 72, 72, 72]
90
Contig2 ...GAATTACTAT...
[68, 68, 68, 68, 68, 68, 68, 68, 68, 68]
90
"""
for ace_contig in Ace.parse(handle):
#Convert the ACE contig record into a SeqRecord...
consensus_seq_str = ace_contig.sequence
#Assume its DNA unless there is a U in it,
if "U" in consensus_seq_str:
if "T" in consensus_seq_str:
#Very odd! Error?
alpha = generic_nucleotide
else:
alpha = generic_rna
else:
alpha = generic_dna
if "*" in consensus_seq_str:
#For consistency with most other file formats, map
#any * gaps into - gaps.
assert "-" not in consensus_seq_str
consensus_seq = Seq(consensus_seq_str.replace("*", "-"),
Gapped(alpha, gap_char="-"))
else:
consensus_seq = Seq(consensus_seq_str, alpha)
#TODO? - Base segments (BS lines) which indicates which read
#phrap has chosen to be the consensus at a particular position.
#Perhaps as SeqFeature objects?
#TODO - Supporting reads (RD lines, plus perhaps QA and DS lines)
#Perhaps as SeqFeature objects?
seq_record = SeqRecord(consensus_seq,
id=ace_contig.name,
name=ace_contig.name)
#Consensus base quality (BQ lines). Note that any gaps (originally
#as * characters) in the consensus do not get a quality entry, so
#we assign a quality of None (zero would be missleading as there may
#be excelent support for having a gap here).
quals = []
i = 0
for base in consensus_seq:
if base == "-":
quals.append(0)
else:
quals.append(ace_contig.quality[i])
i += 1
assert i == len(ace_contig.quality)
seq_record.letter_annotations["phred_quality"] = quals
yield seq_record
def AceIterator(source):
"""Return SeqRecord objects from an ACE file.
This uses the Bio.Sequencing.Ace module to do the hard work. Note that
by iterating over the file in a single pass, we are forced to ignore any
WA, CT, RT or WR footer tags.
Ace files include the base quality for each position, which are taken
to be PHRED style scores. Just as if you had read in a FASTQ or QUAL file
using PHRED scores using Bio.SeqIO, these are stored in the SeqRecord's
letter_annotations dictionary under the "phred_quality" key.
>>> from Bio import SeqIO
>>> with open("Ace/consed_sample.ace") as handle:
... for record in SeqIO.parse(handle, "ace"):
... print("%s %s... %i" % (record.id, record.seq[:10], len(record)))
... print(max(record.letter_annotations["phred_quality"]))
Contig1 agccccgggc... 1475
90
However, ACE files do not include a base quality for any gaps in the
consensus sequence, and these are represented in Biopython with a quality
of zero. Using zero is perhaps misleading as there may be very strong
evidence to support the gap in the consensus. Previous versions of
Biopython therefore used None instead, but this complicated usage, and
prevented output of the gapped sequence as FASTQ format.
>>> from Bio import SeqIO
>>> with open("Ace/contig1.ace") as handle:
... for record in SeqIO.parse(handle, "ace"):
... print("%s ...%s..." % (record.id, record.seq[85:95]))
... print(record.letter_annotations["phred_quality"][85:95])
... print(max(record.letter_annotations["phred_quality"]))
Contig1 ...AGAGG-ATGC...
[57, 57, 54, 57, 57, 0, 57, 72, 72, 72]
90
Contig2 ...GAATTACTAT...
[68, 68, 68, 68, 68, 68, 68, 68, 68, 68]
90
"""
for ace_contig in Ace.parse(source):
# Convert the ACE contig record into a SeqRecord...
consensus_seq_str = ace_contig.sequence
# Assume its DNA unless there is a U in it,
if "U" in consensus_seq_str:
if "T" in consensus_seq_str:
# Very odd! Error?
alpha = generic_nucleotide
else:
alpha = generic_rna
else:
alpha = generic_dna
if "*" in consensus_seq_str:
# For consistency with most other file formats, map
# any * gaps into - gaps.
assert "-" not in consensus_seq_str
consensus_seq = Seq(consensus_seq_str.replace("*", "-"), alpha)
else:
consensus_seq = Seq(consensus_seq_str, alpha)
# TODO? - Base segments (BS lines) which indicates which read
# phrap has chosen to be the consensus at a particular position.
# Perhaps as SeqFeature objects?
# TODO - Supporting reads (RD lines, plus perhaps QA and DS lines)
# Perhaps as SeqFeature objects?
seq_record = SeqRecord(consensus_seq,
id=ace_contig.name,
name=ace_contig.name)
# Consensus base quality (BQ lines). Note that any gaps (originally
# as * characters) in the consensus do not get a quality entry, so
# we assign a quality of None (zero would be misleading as there may
# be excellent support for having a gap here).
quals = []
i = 0
for base in consensus_seq:
if base == "-":
quals.append(0)
else:
quals.append(ace_contig.quality[i])
i += 1
assert i == len(ace_contig.quality)
seq_record.letter_annotations["phred_quality"] = quals
yield seq_record
def snp_count(in_ace, out_file, snp_dict, tags, win_len, max_del, stars):
"""Genotype individuals at SNPs loci.
"""
win_buffer = (win_len - 1) / 2
ace_gen = Ace.parse(open(in_ace, 'r'))
with open(out_file, "w") as output_file:
output_file.write("Contig_nb\tPos\ttag_name\tA\tC\tG\tT\tN\t*\t-\n")
while 1:
try:
contig = ace_gen.next()
except:
print "***All contigs treated***"
break
align = Alignment(Gapped(IUPAC.ambiguous_dna, "-"))
align.add_sequence(contig.name, contig.sequence)
for readn in xrange(len(contig.reads)):
clipst = contig.reads[readn].qa.qual_clipping_start # GOOD
clipe = contig.reads[readn].qa.qual_clipping_end # GOOD
clipst2 = contig.reads[readn].qa.align_clipping_start # Added
clipe2 = contig.reads[readn].qa.align_clipping_end # Added
if clipst2 > clipst: # Added
clipst = clipst2 # Added
if clipe2 < clipe2: # Added
clipe = clipe2 # Added
start = contig.af[readn].padded_start
seq = cut_ends(contig.reads[readn].rd.sequence, clipst, clipe)
seq = pad_read(seq, start, len(contig.sequence))
if "pseudo" not in contig.reads[readn].rd.name:
align.add_sequence(contig.reads[readn].rd.name, seq)
sequences = read_fasta(align.format("fasta"))
contig_name = re.findall("(Contig_[0-9]+)", sequences[0][0])[0]
print "Treating", contig_name
positions = []
try:
positions = snp_dict[contig_name]
except:
continue
d = {}
for pos in positions:
if stars == True:
pos_ok = correct_position(pos, sequences[0][1])
else:
pos_ok = pos
left = pos_ok - 5
if left < 0:
left = 0
right = pos_ok + 1 + 5 # takes into account the middle nucleotide
ref_window = sequences[0][1][left:right]
d.setdefault(pos, {})
d[pos].setdefault("XX_noTag", {})
for nuc in list("ACGTN*-"):
d[pos]["XX_noTag"].setdefault(nuc, 0)
for tag in tags:
d[pos].setdefault(tag, {})
for nuc in list("ACGTN*-"):
d[pos][tag].setdefault(nuc, 0)
for fasta in sequences:
window = fasta[1][left:right]
del_count = 0
if window.count("-") > win_buffer - 3:
continue # Need at least 3 nucleotides on each side
for tag in tags:
if tag in fasta[0]:
t = tag
break
else:
t = "XX_noTag"
if len(ref_window) == len(window):
for i in xrange(len(window)):
if ref_window[i].isalpha() and window[i] == "*" or \
window[i].isalpha() and ref_window[i] == "*":
del_count += 1
if del_count > max_del:
continue
p = pos
s = fasta[1] # Sequence
n = s[pos_ok - 1].upper()
d[p][t][n] += 1
for p in sorted(d):
for t in sorted(d[p]):
output_file.write(contig_name + "\t" + str(p) + "\t" +
str(t))
for n in list("ACGTN*-"):
output_file.write("\t" + str(d[p][t][n]))
output_file.write("\n")
