def write_bam_byChr(bamFile, chr, excludeBamFile, editBamFile, modifyReadsName,
deleteReadsName, addReadsName):
print chr
bam = pysam.AlignmentFile(bamFile, 'rb')
excludeBam = pysam.AlignmentFile(excludeBamFile, 'wb', template=bam)
editBam = pysam.AlignmentFile(editBamFile, 'wb', template=bam)
# print bam
delete = open(editBamFile + ".del", 'w')
m = 0
for read in bam.fetch(chr):
# print read
m += 1
keyname = getKeyName(read)
if keyname in modifyReadsName:
editBam.write(read)
elif keyname in deleteReadsName:
delete.write(keyname + "\n")
continue
elif keyname in addReadsName:
editBam.write(read)
excludeBam.write(read)
else:
excludeBam.write(read)
delete.close()
print "Total reads: ", m
bam.close()
excludeBam.close()
editBam.close()
def write_bam_byChr(bamFile, chr, excludeBamFile, editBamFile,
total_modify_reads_file, total_delete_reads_file,
total_add_reads_file):
print chr
bam = pysam.AlignmentFile(bamFile, 'rb')
excludeBam = pysam.AlignmentFile(excludeBamFile, 'wb', template=bam)
editBam = pysam.AlignmentFile(editBamFile, 'wb', template=bam)
delete = open(editBamFile + ".del", 'w')
m = 0
modifyReadsName = get_name_list(total_modify_reads_file, chr)
deleteReadsName = get_name_list(total_delete_reads_file, chr)
addReadsName = get_name_list(total_add_reads_file, chr)
for read in bam.fetch(chr):
m += 1
keyname = getKeyName(read)
if keyname in modifyReadsName:
editBam.write(read)
elif keyname in deleteReadsName:
delete.write(keyname + "\n")
continue
elif keyname in addReadsName:
editBam.write(read)
excludeBam.write(read)
else:
excludeBam.write(read)
delete.close()
print "Total reads: ", m
bam.close()
excludeBam.close()
editBam.close()
def _get_write(total_reads, reads_file_out, reads_pair_out):
reads_file = open(reads_file_out, 'w')
reads_pair = open(reads_pair_out, 'w')
for read_pair in total_reads:
tmp = []
for read in read_pair:
read_name = getKeyName(read)
tmp.append(read_name)
reads_file.write(str(read))
reads_pair.write("%s\n" % ",".join(tmp))
reads_file.close()
reads_pair.close()
def merge_edit_reads(total_modify_reads, total_add_reads, total_delete_reads):
total_modify_readname_list, total_add_readname_list, total_delete_readname_list = [], [], []
total_modify_reads_list, total_add_reads_list, total_delete_reads_list = [], [], []
for read_pair in total_modify_reads:
for read in read_pair:
read_name = getKeyName(read)
total_modify_readname_list.append(read_name)
total_modify_reads_list.append(read)
for read_pair in total_add_reads:
for read in read_pair:
read_name = getKeyName(read)
total_add_readname_list.append(read_name)
total_add_reads_list.append(read)
for read_pair in total_delete_reads:
for read in read_pair:
read_name = getKeyName(read)
total_delete_readname_list.append(read_name)
total_delete_reads_list.append(read)
return total_modify_readname_list, total_delete_readname_list, total_add_readname_list
def get_write_reads(total_modify_reads, total_delete_reads, total_add_reads,
total_reads_file_dict, total_reads_list_dict):
for typ, reads_dict in zip(
('modify', 'delete', 'add'),
(total_modify_reads, total_delete_reads, total_add_reads)):
reads_file = total_reads_file_dict[typ]
reads_pair = total_reads_list_dict[typ]
for read_pair in reads_dict:
tmp = []
for read in read_pair:
read_name = getKeyName(read)
tmp.append(read_name)
reads_file.write(str(read))
reads_pair.write("%s\n" % ",".join(tmp))
def write_sub_bam(chrom_list, used_bam_file_tmp, exclude_bam_file_tmp, out_dir,
total_modify_readname_list, total_delete_readname_list,
total_add_readname_list, process):
write_bam_pool = Pool(int(process))
exclude_bam_list = [exclude_bam_file_tmp]
usedBamList = []
for chrom in chrom_list:
excludeBam_chr = "%s/exclude_%s.bam" % (out_dir, chrom)
exclude_bam_list.append(excludeBam_chr)
usedBam_chr = "%s/used_%s.bam" % (out_dir, chrom)
usedBamList.append(usedBam_chr)
write_bam_pool.apply_async(
write_bam_byChr,
args=(used_bam_file_tmp, chrom, excludeBam_chr, usedBam_chr,
total_modify_readname_list, total_delete_readname_list,
total_add_readname_list))
write_bam_pool.close()
write_bam_pool.join()
exclude_bam_file = os.path.join(out_dir, "exclude.bam")
bamMerge(exclude_bam_list, exclude_bam_file)
used_bam_file = os.path.join(out_dir, "used.bam")
if len(usedBamList) != 1:
bamMerge(usedBamList, used_bam_file)
else:
used_bam_file = usedBamList[0]
bamSort(used_bam_file, os.path.join(out_dir, "used.sort"))
used_sort_bam_file = os.path.join(out_dir, "used.sort.bam")
bamIndex(used_sort_bam_file)
used_bam = pysam.AlignmentFile(used_sort_bam_file, 'rb')
used_reads = {}
for read in used_bam.fetch():
keyname = getKeyName(read)
used_reads[keyname] = read
used_bam.close()
return used_reads, used_bam_file, exclude_bam_file
def merge_edit_bam(bam_file, out_dir, is_single, total_modify_reads,
total_add_reads, used_reads, seqer, aligner, aligner_index,
flow_order, lib_key, barcode, tag):
bam = pysam.AlignmentFile(bam_file, 'rb')
edit_bam_file = os.path.join(out_dir, "edit.bam")
edit_bam = pysam.AlignmentFile(edit_bam_file, 'wb', template=bam)
readname_convert_file = os.path.join(out_dir, "readname_convert.txt")
fout_convert = open(readname_convert_file, 'w')
edit_bam_reads = {}
if is_single:
for read_pair in total_modify_reads:
read1 = read_pair[0]
keyname_read1 = getKeyName(read1)
orig_read1 = used_reads[keyname_read1]
new_read1 = copy.deepcopy(orig_read1)
new_read1.query_sequence = read1.query_sequence
new_read1.query_qualities = read1.query_qualities
new_name = read1.query_name.split(
":")[0] + ":" + get_new_readname()
new_read1.query_name = new_name
if seqer == "life":
new_read1 = deal_life_reads(new_read1, flow_order, lib_key,
barcode)
if tag:
new_read1 = add_tag(new_read1)
edit_bam.write(new_read1)
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read1.query_name, new_read1.is_read1,
new_read1.reference_start, new_read1.reference_end))
strand = getReadStrand(new_read1)
if new_name not in edit_bam_reads:
edit_bam_reads[new_name] = dict()
edit_bam_reads[new_name][strand] = new_read1
for read_pair in total_add_reads:
read1 = read_pair[0]
keyname_read1 = getKeyName(read1)
orig_read1 = used_reads[keyname_read1]
new_read1 = copy.deepcopy(orig_read1)
new_name = get_new_readname()
new_read1.query_name = new_name
edit_bam.write(new_read1)
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read1.query_name, new_read1.is_read1,
new_read1.reference_start, new_read1.reference_end))
strand = getReadStrand(new_read1)
if new_name not in edit_bam_reads:
edit_bam_reads[new_name] = dict()
edit_bam_reads[new_name][strand] = new_read1
else:
for read_pair in total_modify_reads + total_add_reads:
read1 = read_pair[0]
read2 = read_pair[1]
keyname_read1 = getKeyName(read1)
keyname_read2 = getKeyName(read2)
orig_read1 = used_reads[keyname_read1]
orig_read2 = used_reads[keyname_read2]
orig_read1_name = orig_read1.query_name
orig_read2_name = orig_read2.query_name
new_read1 = copy.deepcopy(orig_read1)
new_read2 = copy.deepcopy(orig_read2)
new_read1.query_sequence = read1.query_sequence
new_read1.query_qualities = read1.query_qualities
new_read2.query_sequence = read2.query_sequence
new_read2.query_qualities = read2.query_qualities
new_name = get_new_readname()
new_read1.query_name = new_name
new_read2.query_name = new_name
strand1 = getReadStrand(new_read1)
strand2 = getReadStrand(new_read2)
if new_name not in edit_bam_reads:
edit_bam_reads[new_name] = dict()
edit_bam_reads[new_name][strand1] = new_read1
edit_bam_reads[new_name][strand2] = new_read2
if tag:
new_read1 = add_tag(new_read1)
new_read2 = add_tag(new_read2)
fout_convert.write("%s: %s, %s, %s, %s, %s-%s, %s-%s\n" % (
new_name,
orig_read1_name,
orig_read2_name,
new_read1.is_read1,
new_read2.is_read2,
new_read1.reference_start,
new_read1.reference_end,
new_read2.reference_start,
new_read2.reference_end,
))
edit_bam.write(new_read1)
edit_bam.write(new_read2)
fout_convert.close()
edit_bam.close()
edit_remap_bam_file = os.path.join(out_dir, "edit.remap.bam")
remap(aligner_index, edit_bam_file, edit_remap_bam_file, aligner,
is_single)
if not is_single:
editRemap = pysam.AlignmentFile(edit_remap_bam_file, 'rb')
editRemapBam_addRG_File = os.path.join(out_dir, "edit.remap.addRG.bam")
bamAddRG(editRemap, edit_bam_reads, bam, editRemapBam_addRG_File)
editRemap.close()
else:
editRemapBam_addRG_File = edit_remap_bam_file
bamIndex(editRemapBam_addRG_File)
return editRemapBam_addRG_File
def merge_edit_bam(bam_file, out_dir, is_single, total_modify_reads_file,
total_add_reads_file, used_bam_file, total_modify_list,
total_add_list, seqer, aligner, aligner_index, flow_order,
lib_key, barcode, tag):
bam = pysam.AlignmentFile(bam_file, 'rb')
edit_bam_file = os.path.join(out_dir, "edit.bam")
edit_bam = pysam.AlignmentFile(edit_bam_file, 'wb', template=bam)
readname_convert_file = os.path.join(out_dir, "readname_convert.txt")
fout_convert = open(readname_convert_file, 'w')
# edit_bam_reads = {}
used_bam = pysam.AlignmentFile(used_bam_file, 'rb')
used_reads = {}
for read in used_bam.fetch():
keyname = getKeyName(read)
used_reads[keyname] = read
used_bam.close()
# modify_read_name_dict = get_newname_dict(total_modify_list)
# add_read_name_dict = get_newname_dict(total_add_list)
modify_reads_seq, modify_reads_quan = get_sequence_dict(
total_modify_reads_file)
add_reads_seq, add_reads_quan = get_sequence_dict(total_add_reads_file)
if is_single:
with open(total_modify_list) as fin:
for line in fin:
if not line:
break
data = line.strip().split(",")
read1_name = data[0]
if read1_name not in used_reads:
continue
orig_read1 = used_reads[read1_name]
new_read1 = copy.deepcopy(orig_read1)
new_read1.query_sequence = modify_reads_seq[read1_name]
new_read1.query_qualities = modify_reads_quan[read1_name]
new_name = get_new_readname()
new_read1.query_name = new_name
if seqer == "life":
new_read1 = deal_life_reads(new_read1, flow_order, lib_key,
barcode)
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read1.query_name, new_read1.is_read1,
new_read1.reference_start, new_read1.reference_end))
edit_bam.write(new_read1)
fin.close()
with open(total_add_list) as fin:
for line in fin:
if not line:
break
data = line.strip().split(",")
read1_name = data[0]
if read1_name not in used_reads:
continue
orig_read1 = used_reads[read1_name]
new_read1 = copy.deepcopy(orig_read1)
new_read1.query_sequence = add_reads_seq[read1_name]
new_read1.query_qualities = add_reads_quan[read1_name]
new_name = get_new_readname()
new_read1.query_name = new_name
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read1.query_name, new_read1.is_read1,
new_read1.reference_start, new_read1.reference_end))
edit_bam.write(new_read1)
fin.close()
else:
with open(total_modify_list) as fin:
for line in fin:
if not line:
break
data = line.strip().split(",")
read1_name, read2_name = data[0], data[1]
if read1_name not in used_reads or read2_name not in used_reads:
continue
orig_read1 = used_reads[read1_name]
orig_read2 = used_reads[read2_name]
new_read1 = copy.deepcopy(orig_read1)
new_read2 = copy.deepcopy(orig_read2)
print read1_name
new_read1.query_sequence = modify_reads_seq[read1_name]
new_read1.query_qualities = modify_reads_quan[read1_name]
new_read2.query_sequence = modify_reads_seq[read2_name]
new_read2.query_qualities = modify_reads_quan[read2_name]
new_name = get_new_readname()
new_read1.query_name = new_name
new_read2.query_name = new_name
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read1.query_name, new_read1.is_read1,
new_read1.reference_start, new_read1.reference_end))
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read2.query_name, new_read2.is_read1,
new_read2.reference_start, new_read2.reference_end))
edit_bam.write(new_read1)
edit_bam.write(new_read2)
fin.close()
with open(total_add_list) as fin:
for line in fin:
if not line:
break
data = line.strip().split(",")
read1_name, read2_name = data[0], data[1]
if read1_name not in used_reads or read2_name not in used_reads:
continue
orig_read1 = used_reads[read1_name]
orig_read2 = used_reads[read2_name]
new_read1 = copy.deepcopy(orig_read1)
new_read2 = copy.deepcopy(orig_read2)
new_read1.query_sequence = add_reads_seq[read1_name]
new_read1.query_qualities = add_reads_quan[read1_name]
new_read2.query_sequence = add_reads_seq[read2_name]
new_read2.query_qualities = add_reads_quan[read2_name]
new_name = get_new_readname()
new_read1.query_name = new_name
new_read2.query_name = new_name
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read1.query_name, new_read1.is_read1,
new_read1.reference_start, new_read1.reference_end))
fout_convert.write(
"%s: %s, %s, %s-%s\n" %
(new_name, orig_read2.query_name, new_read2.is_read1,
new_read2.reference_start, new_read2.reference_end))
edit_bam.write(new_read1)
edit_bam.write(new_read2)
fin.close()
edit_bam.close()
header = os.path.join(out_dir, 'bam.header')
os.system('samtools view -H %s|grep "^@RG" > %s' % (bam_file, header))
head = open(header, 'r').readline().rstrip()
if not head:
head = None
edit_remap_bam_file = os.path.join(out_dir, "edit.remap.bam")
remap(aligner_index, edit_bam_file, edit_remap_bam_file, aligner,
is_single, head)
edit_remap_bam_sorted_prefix = os.path.join(out_dir, "edit.remap.sort")
edit_remap_bam_sorted_file = os.path.join(out_dir, "edit.remap.sort.bam")
bamSort(edit_remap_bam_file, edit_remap_bam_sorted_prefix)
return edit_remap_bam_sorted_file