def __generate_test_data__(self, bam1, sam2, contig_positions, min_mapq):
run_results = {
"total_queries": 0,
"passed_filter": 0,
"date": datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),
"min_mapq": min_mapq,
"query_sam": sam1,
"searched_sam": sam2,
"contig_positions": contig_positions,
"pickle": None,
}
contig_2_contig_dict = defaultdict(list)
# CREATE SAM INDEX IF IT DOESN'T EXIST
if not os.path.exists(sam2 + FileIndex.ext):
FileIndex.create(sam2, lambda fh: SamLine(fh).qname, allow_multiple=True)
fi = FileIndex(sam2, SamLine, allow_multiple=True)
ma = MergeAssemblies()
contig_starts = [int(l.strip()) for l in open(contig_positions, 'rU')]
fq1_out = open("test_fq1.10k.fq", 'w')
fq2_out = open("test_fq2.10k.fq", 'w')
with open(sam1, 'rU') as qs:
for count, q in enumerate(qs):
if q.startswith("@"): # skip header lines
continue
# SPLIT LINE AND IDENTIFY QUERY POSITION.
q = q.strip().split("\t")
query_pos = int(q[3])
query_pos = ma.round_bp_pos(query_pos) # this could be more sophisticated.
q_id = q[0][:-1] + "2" # create query id (e.g., ends with "2")
q_mapq = int(q[4]) # and, mapping quality.
if count > 10000: break
# SEARCH FOR QUERY AND PARSE RESULTS
for s in fi[q_id]:
fq1 = (q[0], q[9], q[10])
fq2 = (s.qname, s.seq, s.qual)
fq1_line = "@{}\n{}\n+\n{}\n".format(*fq1)
fq2_line = "@{}\n{}\n+\n{}\n".format(*fq2)
fq1_out.write(fq1_line)
fq2_out.write(fq2_line)
import os.path as op
sys.path.insert(0, op.join(op.dirname(__file__), ".."))
from fileindex import FileIndex
class FastQEntry(object):
__slots__ = ("name", "seq", "l3", "qual", "fpos")
def __init__(self, fh):
self.name = fh.readline().rstrip("\r\n")
self.seq = fh.readline().rstrip("\r\n")
self.l3 = fh.readline().rstrip("\r\n")
self.qual = fh.readline().rstrip("\r\n")
if __name__ == "__main__":
f = "/usr/local/src/bowtie/bowtie-0.12.1/work/reads/s_1_sequence.txt"
N = 100
# if not op.exists(f + FileIndex.ext):
FileIndex.create(f, lambda fh: FastQEntry(fh).name)
fi = FileIndex(f, FastQEntry)
print "getting %i keys..." % N
for i, k in enumerate(fi.db.iterkeys(str)):
print fi[k].seq
if i == N:
break
import sys
import os.path as op
sys.path.insert(0, op.join(op.dirname(__file__), ".."))
from fileindex import FileIndex
class SamLine(object):
__slots__ = ('name', 'ref_loc', 'ref_seqid')
def __init__(self, fh):
line = fh.readline().split("\t") or [None]
self.name = line[0]
self.ref_seqid = line[2]
self.ref_loc = int(line[3])
# ... other sam format stuff omitted.
if __name__ == "__main__":
f = '/usr/local/src/methylcode/emen/en-data/out/methylcoded.sam'
if not op.exists(f + FileIndex.ext):
FileIndex.create(f, lambda fh: SamLine(fh).name, allow_multiple=True)
fi = FileIndex(f, SamLine, allow_multiple=True)
print [(s.name, s.ref_seqid, s.ref_loc) for s in fi['23351265']]
def associate_contigs(self, bam1, sam2, min_mapq, run_ID, R1_starts=None):
# print bam1, sam2, min_mapq, run_ID
# sys.exit()
run_results = {
"total_queries": 0,
"passed_filter": 0,
"date": datetime.datetime.now().strftime("%Y-%m-%d %H:%M"),
"min_mapq": min_mapq,
"query_sam": bam1,
"searched_sam": sam2,
# "contig_positions": contig_positions,
"pickle": None,
"failed_filter":0,
}
contig_2_contig_dict = defaultdict(list)
# CREATE SAM INDEX IF IT DOESN'T EXIST
if not os.path.exists(sam2 + FileIndex.ext):
FileIndex.create(sam2, lambda fh: SamLine(fh).qname, allow_multiple=True)
fi = FileIndex(sam2, SamLine, allow_multiple=True)
ma = MergeAssemblies()
R1_sam = pysam.Samfile(bam1,'rb')
path = os.path.split(bam1)[0]
if R1_starts == None:
R1_starts = os.path.join(path, "{}.R1.contig_start_pos.txt".format(run_ID))
low_depth_R1_starts = os.path.join(path, "{}.R1.contig_start_pos.no_pass.txt".format(run_ID))
low_depth_R1_starts = open(low_depth_R1_starts, 'w')
all_R1s = set()
# ITERATE OVER START POSITIONS
for count, start in enumerate(open(R1_starts,'rU')):
query_pos = int(start.strip())
all_R1s.update([query_pos])
if query_pos is 0:
continue
#print "{}.R1".format(run_ID), query_pos-5, query_pos+5
# samtools view alignments/Trachs_Merged.1.fq.sorted.bam Trachs_Merged_stacks.R1:590-600
reads = R1_sam.fetch("{}.R1".format(run_ID), query_pos-5, query_pos+5)
reads = [r for r in reads] # unpack interator
depth = len(reads)
# if depth < min_depth:
# unassociated_R1s.update([query_pos])
for q in reads:
q_id = q.qname[:-1] + "2" # create query id (e.g., ends with "2")
q_mapq = q.mapq # and, mapping quality.
# SEARCH FOR QUERY AND PARSE RESULTS
for s in fi[q_id]:
if (q_mapq > min_mapq) and (s.mapq > min_mapq):
hit_pos = ma.get_hit_pos(s)
hit_pos = ma.round_bp_pos(hit_pos)
contig_2_contig_dict[query_pos].append(hit_pos)
run_results["passed_filter"] = len(contig_2_contig_dict.keys())
unassociated_R1s = all_R1s.difference(set(contig_2_contig_dict.keys()))
run_results["failed_filter"] = len(unassociated_R1s)
[low_depth_R1_starts.write("{}\n".format(p)) for p in unassociated_R1s]
#CREATE PICKLE OUTPUT FILE
today = datetime.date.today()
path = os.path.split(bam1)[0]
pkl_output_file_name = os.path.join(path, '{}.R1_to_R2_contig_associations.pkl'.format(run_ID))
pkl_out = open(pkl_output_file_name, 'wb')
pickle.dump(contig_2_contig_dict, pkl_out)
low_depth_R1_starts.close()
# UPDATE RUN RESULTS AND GENERATE LOGFILE
run_results["total_queries"] = count
run_results['pickle'] = pkl_output_file_name
self.__associate_contigs_log__(run_results, path) # format logging results
return 1
import sys
import os.path as op
sys.path.insert(0, op.join(op.dirname(__file__), ".."))
from fileindex import FileIndex
class FastQEntry(object):
__slots__ = ('name', 'seq', 'l3', 'qual', 'fpos')
def __init__(self, fh):
self.name = fh.readline().rstrip('\r\n')
self.seq = fh.readline().rstrip('\r\n')
self.l3 = fh.readline().rstrip('\r\n')
self.qual = fh.readline().rstrip('\r\n')
if __name__ == "__main__":
f = '/usr/local/src/bowtie/bowtie-0.12.1/work/reads/s_1_sequence.txt'
N = 100
#if not op.exists(f + FileIndex.ext):
FileIndex.create(f, lambda fh: FastQEntry(fh).name)
fi = FileIndex(f, FastQEntry)
print "getting %i keys..." % N
for i, k in enumerate(fi.db.iterkeys(str)):
print fi[k].seq
if i == N: break