def mauve_pw_align(ref, query, dirs):
"""Set up and perform a pairwise alignment with Mauve."""
# set outputs
mauve_outfile = dirs['mauve']+ref.name+"_"+query.name+".mauve"
segfile = dirs['aln_segs']+ref.name+"_"+query.name+"_segs.txt"
# check for existing alignment
if path.exists(segfile):
print "already done"
else:
# prep segments file
open(segfile, 'w').write('')
# purge any pre-existing sslist files
sslist_files = from_dir(dirs['seqfiles'], re.compile(r'.*\.sslist.*'))
for sslist in sslist_files:
try: os.remove(dirs['seqfiles']+sslist)
except Exception: raise
# do Mauve alignment
file_list = [ref.gbk, query.gbk]
align_mauve(file_list, mauve_outfile)
try:
# parse Mauve output (without initial clumping)
coords = mauver_load2_k0(mauve_outfile+".backbone", 0)
print "\nSegment results:", len(coords), '->',
# chop segments that are too long
chop_array = chop_rows(coords, max_size, chop_mode)
print len(chop_array), 'segments <', max_size, 'bp'
# make detailed pairwise alignments of the segments
print "Aligning segments ..."
ref_rec = load_genbank(ref.gbk)
query_rec = load_genbank(query.gbk)
id = iter_align(chop_array, ref_rec, query_rec,
dirs['aln_segs'], segfile)
print "Results:", id, "% id. overall"
except IOError:
print "\nERROR: Mauve alignment failed"
def mauve_pw_align(ref, query, r_root_dir, g_root_dir, dirs, run, max_size,
chop_mode, mauve_exec, mtype):
"""Set up and perform a pairwise alignment with Mauve."""
aln_dir = r_root_dir + run + dirs['aln_segs']
mauve_dir = r_root_dir + run + dirs['mauve']
# set outputs
mauve_outfile = mauve_dir + ref.name + "_" + query.name + ".mauve"
segfile = aln_dir + ref.name + "_" + query.name + "_segs.txt"
# check for existing alignment
if path.exists(segfile):
print "already done"
else:
# prep segments file
open(segfile, 'w').write('')
# purge any pre-existing sslist files
sslist_files = from_dir(g_root_dir, re.compile(r'.*\.sslist.*'))
for sslist in sslist_files:
try:
os.remove(g_root_dir + sslist)
except Exception:
raise
# do Mauve alignment
file_list = [ref.gbk, query.gbk]
align_mauve(file_list, mauve_outfile, mauve_exec)
try:
# parse Mauve output (without initial clumping)
coords = mauver_load2_k0(mauve_outfile + ".backbone", 0, mtype)
print "\nSegment results:", len(coords), '->',
# chop segments that are too long
chop_array = chop_rows(coords, max_size, chop_mode, mtype)
print len(chop_array), 'segments <', max_size, 'bp'
# make detailed pairwise alignments of the segments
print "Aligning segments ..."
ref_rec = load_genbank(ref.gbk)
query_rec = load_genbank(query.gbk)
id = iter_align(chop_array, ref_rec, query_rec, aln_dir, segfile)
print "Results:", id, "% id. overall"
except IOError:
print "\nERROR: Mauve alignment failed"
raise
def build_scaffolds(run_ref, r_root_dir, run_dirs, prox_D, separator, genomes,
run_id, timestamp, mtype, mode):
"""Build a scaffold of contigs based on the reference.
This takes contigs that gave positive hits when blasted with reference
segments. The contigs were aligned against the complete reference in a
previous step for mapping purposes. Now the output of that step is re-used
determine their position. A caveat is that if there are natural local
rearrangements in the sequence relative to the reference, they may not be
resolved appropriately. The problem is somewhat moderated by the fact that
this function takes the best (usually the largest) hit region as "anchor"
to position the contig within the scaffold. But if the rearranged region
takes up a significant portion of the contig length, the anchoring will
probably not be called correctly. Visual inspection of the finalized
maps should help diagnose any such problems. The order can be fixed
manually using the Mauve Contig Mover, which is part of Mauve 2.
Note that not all hit contigs are "real" hits, so filtering should be
applied before scaffolding to generate constructs.
Model-based filtering produces a list of contigs that will be passed to
the scaffolder. If filtering manually by looking at the maps,
there are two options available: either select exclusively OR exclude a
subset of contigs for the scaffolding process. This is done by listing
their ID number in the genome dictionaries in the config file then
resuming the pipeline from this step.
"""
# set inputs and outputs
ref_n = run_ref.name
run_root = r_root_dir + run_id + "/"
ctgs_root = run_root + run_dirs['run_gbk_ctgs_dir'] + ref_n + "/"
mauve_root = run_root + run_dirs['mauve_out_dir'] + ref_n + "/contigs/"
scaffolds_dir = run_root + run_dirs['scaffolds_dir'] + ref_n + "/"
print " ", ref_n
# log
logstring = "".join(["\n\n# Build scaffold constructs @", timestamp, "\n"])
run_ref.log(logstring)
# cycle through genomes
for genome in genomes:
# set inputs
g_name = genome['name']
ctgs_dir = ctgs_root + g_name + "/"
print "\t", g_name, "...",
# log
logstring = "".join(["\n", g_name])
run_ref.log(logstring)
# set outputs
mauve_dir = mauve_root + g_name + "/"
ensure_dir([mauve_dir, scaffolds_dir])
scaff_fas = scaffolds_dir + g_name + "_" + ref_n + "_scaffold.fas"
scaff_gbk = scaffolds_dir + g_name + "_" + ref_n + "_scaffold.gbk"
# list genbank files in matches directory
dir_contents = listdir(ctgs_dir)
anchors_array = np.zeros(1,
dtype=[('ctg', 'i4'), ('start', 'i4'),
('end', 'i4'), ('orient', 'i2')])
# identify contigs we want to select
subset = []
for item in dir_contents:
pattern = re.compile(r'.*_(\d*)\.gbk$')
match = pattern.match(item)
if match:
ctg_num = match.group(1)
if mode == "exclude":
try:
if int(ctg_num) in genome[mode]:
msg = "(" + ctg_num + ")"
print msg,
run_ref.log(msg)
else:
subset.append(ctg_num)
except KeyError:
msg = "WARNING: no ignored segments list, including all"
print msg
msg = ctg_num
print msg,
subset.append(ctg_num)
run_ref.log(msg)
elif mode == "select":
try:
if int(ctg_num) in genome[mode]:
msg = ctg_num
print msg,
run_ref.log(msg)
subset.append(ctg_num)
else:
msg = "(" + ctg_num + ")"
print msg,
run_ref.log(msg)
except KeyError:
msg = "WARNING: no selected segments list, including all"
print msg
msg = ctg_num
print msg,
subset.append(ctg_num)
run_ref.log(msg)
# at this point we should have a subset of contigs selected
for ctg_num in subset:
logstring = "".join(["\t", ctg_num])
run_ref.log(logstring)
# set inputs
mauve_file = mauve_dir + ctg_num + ".mauve"
bb_file = mauve_file + ".backbone"
try:
# parse Mauve output
coords = mauver_load2_k0(bb_file, prox_D, mtype)
# determine which segment to use as anchor
anchor_seg = get_anchor_loc(coords)
anchors_array = np.insert(
anchors_array, 0,
(ctg_num, anchor_seg['start'], anchor_seg['end'],
anchor_seg['orient']))
except IOError:
msg = "\tERROR: Mauve alignment not found\n\t"
print msg
run_ref.log(msg)
except Exception:
msg = "\tERROR: Iteration failure\n\t"
print msg
run_ref.log(msg)
# abort if there is no valid contig to proceed with
try:
assert len(anchors_array) > 1 # always 1 left from stub
except AssertionError:
msg = "\tWARNING: Contig list empty\n\t"
print msg
run_ref.log(msg)
else:
# order contigs by anchor location
anchors_array = np.sort(anchors_array, order='start')
# load contig records from the genbank files in the matches directory
ctg_list = []
for ctg_anchor in anchors_array:
ctg_num = ctg_anchor['ctg']
if ctg_num > 0:
contig_gbk = ctgs_dir + g_name + "_" + str(
ctg_num) + ".gbk"
record = load_genbank(contig_gbk)
if ctg_anchor['orient'] == -1: # flip record
record = record.reverse_complement(id=True,
name=True,
annotations=True,
description=True)
ctg_list.append(record)
else: # workaround for having 0 value leftover from stub
pass # having it might come in handy in later dev
# output scaffold files
write_fasta(scaff_fas, ctg_list)
scaff_record = SeqRecord('', id='temp')
scaff_bumper = SeqRecord(separator, id='join')
for record in ctg_list:
feat_start = len(scaff_record.seq)
scaff_record += record
feat_stop = len(scaff_record.seq)
scaff_record += scaff_bumper
feat_loc = FeatureLocation(feat_start, feat_stop)
pattern = re.compile(r'.*_(\d*)$')
match = pattern.match(record.id)
try:
ctg_num = match.group(1)
except Exception:
ctg_num = 'N'
feature = SeqFeature(location=feat_loc,
type='contig',
qualifiers={'id': ctg_num})
scaff_record.features.append(feature)
scaff_record.description = g_name + " scaffold from " + ref_n
try:
scaff_record.id = g_name
write_genbank(scaff_gbk, scaff_record[:-100]) # rm last bumper
except ValueError:
scaff_record.id = g_name[:10]
write_genbank(scaff_gbk, scaff_record[:-100]) # rm last bumper
print ""
def build_scaffolds(run_ref, r_root_dir, run_dirs, prox_D, separator,
genomes, run_id, timestamp, mtype, mode):
"""Build a scaffold of contigs based on the reference.
This takes contigs that gave positive hits when blasted with reference
segments. The contigs were aligned against the complete reference in a
previous step for mapping purposes. Now the output of that step is re-used
determine their position. A caveat is that if there are natural local
rearrangements in the sequence relative to the reference, they may not be
resolved appropriately. The problem is somewhat moderated by the fact that
this function takes the best (usually the largest) hit region as "anchor"
to position the contig within the scaffold. But if the rearranged region
takes up a significant portion of the contig length, the anchoring will
probably not be called correctly. Visual inspection of the finalized
maps should help diagnose any such problems. The order can be fixed
manually using the Mauve Contig Mover, which is part of Mauve 2.
Note that not all hit contigs are "real" hits, so filtering should be
applied before scaffolding to generate constructs.
Model-based filtering produces a list of contigs that will be passed to
the scaffolder. If filtering manually by looking at the maps,
there are two options available: either select exclusively OR exclude a
subset of contigs for the scaffolding process. This is done by listing
their ID number in the genome dictionaries in the config file then
resuming the pipeline from this step.
"""
# set inputs and outputs
ref_n = run_ref.name
run_root = r_root_dir+run_id+"/"
ctgs_root = run_root+run_dirs['run_gbk_ctgs_dir']+ref_n+"/"
mauve_root = run_root+run_dirs['mauve_out_dir']+ref_n+"/contigs/"
scaffolds_dir = run_root+run_dirs['scaffolds_dir']+ref_n+"/"
print " ", ref_n
# log
logstring = "".join(["\n\n# Build scaffold constructs @", timestamp, "\n"])
run_ref.log(logstring)
# cycle through genomes
for genome in genomes:
# set inputs
g_name = genome['name']
ctgs_dir = ctgs_root+g_name+"/"
print "\t", g_name, "...",
# log
logstring = "".join(["\n", g_name])
run_ref.log(logstring)
# set outputs
mauve_dir = mauve_root+g_name+"/"
ensure_dir([mauve_dir, scaffolds_dir])
scaff_fas = scaffolds_dir+g_name+"_"+ref_n+"_scaffold.fas"
scaff_gbk = scaffolds_dir+g_name+"_"+ref_n+"_scaffold.gbk"
# list genbank files in matches directory
dir_contents = listdir(ctgs_dir)
anchors_array = np.zeros(1, dtype=[('ctg', 'i4'),
('start', 'i4'),
('end', 'i4'),
('orient', 'i2')])
# identify contigs we want to select
subset = []
for item in dir_contents:
pattern = re.compile(r'.*_(\d*)\.gbk$')
match = pattern.match(item)
if match:
ctg_num = match.group(1)
if mode == "exclude":
try:
if int(ctg_num) in genome[mode]:
msg = "("+ctg_num+")"
print msg,
run_ref.log(msg)
else:
subset.append(ctg_num)
except KeyError:
msg = "WARNING: no ignored segments list, including all"
print msg
msg = ctg_num
print msg,
subset.append(ctg_num)
run_ref.log(msg)
elif mode == "select":
try:
if int(ctg_num) in genome[mode]:
msg = ctg_num
print msg,
run_ref.log(msg)
subset.append(ctg_num)
else:
msg = "("+ctg_num+")"
print msg,
run_ref.log(msg)
except KeyError:
msg = "WARNING: no selected segments list, including all"
print msg
msg = ctg_num
print msg,
subset.append(ctg_num)
run_ref.log(msg)
# at this point we should have a subset of contigs selected
for ctg_num in subset:
logstring = "".join(["\t", ctg_num])
run_ref.log(logstring)
# set inputs
mauve_file = mauve_dir+ctg_num+".mauve"
bb_file = mauve_file+".backbone"
try:
# parse Mauve output
coords = mauver_load2_k0(bb_file, prox_D, mtype)
# determine which segment to use as anchor
anchor_seg = get_anchor_loc(coords)
anchors_array = np.insert(anchors_array, 0,
(ctg_num,
anchor_seg['start'],
anchor_seg['end'],
anchor_seg['orient']))
except IOError:
msg = "\tERROR: Mauve alignment not found\n\t"
print msg
run_ref.log(msg)
except Exception:
msg = "\tERROR: Iteration failure\n\t"
print msg
run_ref.log(msg)
# abort if there is no valid contig to proceed with
try:
assert len(anchors_array) > 1 # always 1 left from stub
except AssertionError:
msg = "\tWARNING: Contig list empty\n\t"
print msg
run_ref.log(msg)
else:
# order contigs by anchor location
anchors_array = np.sort(anchors_array, order='start')
# load contig records from the genbank files in the matches directory
ctg_list = []
for ctg_anchor in anchors_array:
ctg_num = ctg_anchor['ctg']
if ctg_num > 0:
contig_gbk = ctgs_dir+g_name+"_"+str(ctg_num)+".gbk"
record = load_genbank(contig_gbk)
if ctg_anchor['orient'] == -1: # flip record
record = record.reverse_complement(id=True, name=True,
annotations=True, description=True)
ctg_list.append(record)
else: # workaround for having 0 value leftover from stub
pass # having it might come in handy in later dev
# output scaffold files
write_fasta(scaff_fas, ctg_list)
scaff_record = SeqRecord('', id='temp')
scaff_bumper = SeqRecord(separator, id='join')
for record in ctg_list:
feat_start = len(scaff_record.seq)
scaff_record += record
feat_stop = len(scaff_record.seq)
scaff_record += scaff_bumper
feat_loc = FeatureLocation(feat_start, feat_stop)
pattern = re.compile(r'.*_(\d*)$')
match = pattern.match(record.id)
try: ctg_num = match.group(1)
except Exception: ctg_num = 'N'
feature = SeqFeature(location=feat_loc,
type='contig',
qualifiers={'id': ctg_num})
scaff_record.features.append(feature)
scaff_record.description = g_name+" scaffold from "+ref_n
try:
scaff_record.id = g_name
write_genbank(scaff_gbk, scaff_record[:-100]) # rm last bumper
except ValueError:
scaff_record.id = g_name[:10]
write_genbank(scaff_gbk, scaff_record[:-100]) # rm last bumper
print ""
def align_ctg2ref(run_ref, run_id, timestamp, r_root_dir, run_dirs, genomes,
mauve_exec, max_size, chop_mode, mtype):
"""Align contigs pairwise to the reference contig."""
# set inputs and outputs
ref_n = run_ref.name
run_root = r_root_dir + run_id + "/"
ref_ctg_file = run_ref.file
mauve_root = run_root + run_dirs['mauve_out_dir'] + ref_n + "/contigs/"
segments_root = run_root + run_dirs['aln_seg_dir'] + ref_n + "/contigs/"
q_ctgs_root = run_root + run_dirs['match_out_dir'] + ref_n + "/"
ensure_dir([segments_root])
print " ", ref_n
# log
logstring = "".join(["\n\n# Align contigs to ref @", timestamp, "\n"])
run_ref.log(logstring)
# cycle through genomes
for genome in genomes:
# set inputs and outputs
g_name = genome['name']
ctgs_fas_dir = q_ctgs_root + g_name + "/"
mauve_dir = mauve_root + g_name + "/"
aln_segs_root = segments_root + g_name + "/"
ensure_dir([mauve_dir])
print "\t", g_name, "...",
# log
logstring = "".join(["\n", g_name])
run_ref.log(logstring)
# list genbank files in matches directory
dir_contents = listdir(ctgs_fas_dir)
for item in dir_contents:
pattern = re.compile(r'.*_(\d*)\.fas$')
match = pattern.match(item)
if match:
ctg_num = match.group(1)
print ctg_num,
logstring = "".join(["\t", ctg_num])
run_ref.log(logstring)
# set inputs and outputs
q_contig = ctgs_fas_dir + item
file_list = (ref_ctg_file, q_contig)
mauve_outfile = mauve_dir + ctg_num + ".mauve"
aln_segs_dir = aln_segs_root + ctg_num + "/"
ensure_dir([aln_segs_dir])
segfile = aln_segs_dir + ctg_num + "_" + ref_n + "_segs.txt"
open(segfile, 'w').write('')
# do Mauve alignment
try:
open(ref_ctg_file, 'r')
open(q_contig, 'r')
except IOError:
msg = "\nERROR: File missing, cannot align\n\t\t\t"
run_ref.log(msg)
print msg
else:
align_mauve(file_list, mauve_outfile, mauve_exec)
try:
# parse Mauve output (without initial clumping)
coords = mauver_load2_k0(mauve_outfile + ".backbone",
0, mtype)
# chop segments that are too long
chop_array = chop_rows(coords, max_size, chop_mode,
mtype)
# make detailed pairwise alignments of the segments
ref_rec = load_genbank(ref_ctg_file)
query_rec = load_fasta(q_contig)
iter_align(chop_array, ref_rec, query_rec,
aln_segs_dir, segfile)
except IOError:
msg = "\nERROR: Mauve alignment failed\n\t\t\t"
run_ref.log(msg)
print msg
except Exception:
msg = "\nERROR: Iteration failed\n\t\t\t"
run_ref.log(msg)
print msg
print ""
def align_cstrct2ref(run_ref, run_id, timestamp, r_root_dir, run_dirs, genomes,
max_size, chop_mode, mtype, mauve_exec):
"""Align constructs pairwise to the reference contig."""
# set inputs and outputs
ref_n = run_ref.name
run_root = r_root_dir + run_id + "/"
ref_ctg_file = run_ref.file
mauve_root = run_root + run_dirs['mauve_out_dir'] + ref_n + "/constructs/"
segments_root = run_root + run_dirs['aln_seg_dir'] + ref_n + "/constructs/"
scaff_root = run_root + run_dirs['scaffolds_dir'] + ref_n + "/"
ensure_dir([segments_root])
print " ", ref_n
# log
logstring = "".join(
["\n\n# Align scaffold constructs to reference @", timestamp, "\n"])
run_ref.log(logstring)
# cycle through genomes
for genome in genomes:
# set inputs
g_name = genome['name']
scaff_gbk = scaff_root + g_name + "_" + ref_n + "_scaffold.gbk"
file_list = (ref_ctg_file, scaff_gbk)
print "\t", g_name, "...",
# log
logstring = "".join(["\n", g_name])
run_ref.log(logstring)
# set outputs
mauve_dir = mauve_root + g_name + "/"
aln_segs_dir = segments_root + g_name + "/"
ensure_dir([mauve_dir, aln_segs_dir])
mauve_outfile = mauve_dir + g_name + "_" + ref_n + ".mauve"
segfile = aln_segs_dir + g_name + "_" + ref_n + "_segs.txt"
# abort if the reference file is not found
try:
open(ref_ctg_file, 'r')
except IOError:
msg = "ERROR: Reference file not found"
print msg
run_ref.log(msg)
raise
# abort if there is no scaffold construct
try:
open(scaff_gbk, 'r')
except IOError:
msg = "WARNING: No scaffold construct to align"
print msg
run_ref.log(msg)
else:
# prep segments file
open(segfile, 'w').write('')
# purge any pre-existing sslist file
sslist_file = scaff_gbk + ".sslist"
if os.path.isfile(sslist_file):
try:
os.remove(sslist_file)
except Exception:
raise
# do Mauve alignment
align_mauve(file_list, mauve_outfile, mauve_exec)
try:
# parse Mauve output (without initial clumping)
coords = mauver_load2_k0(mauve_outfile + ".backbone", 0, mtype)
print len(coords), '->',
logstring = "".join(["\t", str(len(coords))])
run_ref.log(logstring)
# chop segments that are too long
chop_array = chop_rows(coords, max_size, chop_mode, mtype)
print len(chop_array), 'segments <', max_size, 'bp',
logstring = "".join(["\t", str(len(chop_array))])
run_ref.log(logstring)
# make detailed pairwise alignments of the segments
ref_rec = load_genbank(ref_ctg_file)
query_rec = load_genbank(scaff_gbk)
id = iter_align(chop_array, ref_rec, query_rec, aln_segs_dir,
segfile)
print "@", id, "% id. overall"
logstring = "".join(["\t", str(id)])
run_ref.log(logstring)
except IOError:
msg = "\nERROR: Mauve alignment failed"
run_ref.log(msg)
print msg
def align_ctg2ref(run_ref, run_id, timestamp, r_root_dir, run_dirs,
genomes, mauve_exec, max_size, chop_mode, mtype):
"""Align contigs pairwise to the reference contig."""
# set inputs and outputs
ref_n = run_ref.name
run_root = r_root_dir+run_id+"/"
ref_ctg_file = run_ref.file
mauve_root = run_root+run_dirs['mauve_out_dir']+ref_n+"/contigs/"
segments_root = run_root+run_dirs['aln_seg_dir']+ref_n+"/contigs/"
q_ctgs_root = run_root+run_dirs['match_out_dir']+ref_n+"/"
ensure_dir([segments_root])
print " ", ref_n
# log
logstring = "".join(["\n\n# Align contigs to ref @", timestamp, "\n"])
run_ref.log(logstring)
# cycle through genomes
for genome in genomes:
# set inputs and outputs
g_name = genome['name']
ctgs_fas_dir = q_ctgs_root+g_name+"/"
mauve_dir = mauve_root+g_name+"/"
aln_segs_root = segments_root+g_name+"/"
ensure_dir([mauve_dir])
print "\t", g_name, "...",
# log
logstring = "".join(["\n", g_name])
run_ref.log(logstring)
# list genbank files in matches directory
dir_contents = listdir(ctgs_fas_dir)
for item in dir_contents:
pattern = re.compile(r'.*_(\d*)\.fas$')
match = pattern.match(item)
if match:
ctg_num = match.group(1)
print ctg_num,
logstring = "".join(["\t", ctg_num])
run_ref.log(logstring)
# set inputs and outputs
q_contig = ctgs_fas_dir+item
file_list = (ref_ctg_file, q_contig)
mauve_outfile = mauve_dir+ctg_num+".mauve"
aln_segs_dir = aln_segs_root+ctg_num+"/"
ensure_dir([aln_segs_dir])
segfile = aln_segs_dir+ctg_num+"_"+ref_n+"_segs.txt"
open(segfile, 'w').write('')
# do Mauve alignment
try:
open(ref_ctg_file, 'r')
open(q_contig, 'r')
except IOError:
msg = "\nERROR: File missing, cannot align\n\t\t\t"
run_ref.log(msg)
print msg
else:
align_mauve(file_list, mauve_outfile, mauve_exec)
try:
# parse Mauve output (without initial clumping)
coords = mauver_load2_k0(mauve_outfile+".backbone",
0, mtype)
# chop segments that are too long
chop_array = chop_rows(coords, max_size, chop_mode,
mtype)
# make detailed pairwise alignments of the segments
ref_rec = load_genbank(ref_ctg_file)
query_rec = load_fasta(q_contig)
iter_align(chop_array, ref_rec, query_rec,
aln_segs_dir, segfile)
except IOError:
msg = "\nERROR: Mauve alignment failed\n\t\t\t"
run_ref.log(msg)
print msg
except Exception:
msg = "\nERROR: Iteration failed\n\t\t\t"
run_ref.log(msg)
print msg
print ""
def align_cstrct2ref(run_ref, run_id, timestamp, r_root_dir, run_dirs,
genomes, max_size, chop_mode, mtype, mauve_exec):
"""Align constructs pairwise to the reference contig."""
# set inputs and outputs
ref_n = run_ref.name
run_root = r_root_dir+run_id+"/"
ref_ctg_file = run_ref.file
mauve_root = run_root+run_dirs['mauve_out_dir']+ref_n+"/constructs/"
segments_root = run_root+run_dirs['aln_seg_dir']+ref_n+"/constructs/"
scaff_root = run_root+run_dirs['scaffolds_dir']+ref_n+"/"
ensure_dir([segments_root])
print " ", ref_n
# log
logstring = "".join(["\n\n# Align scaffold constructs to reference @",
timestamp, "\n"])
run_ref.log(logstring)
# cycle through genomes
for genome in genomes:
# set inputs
g_name = genome['name']
scaff_gbk = scaff_root+g_name+"_"+ref_n+"_scaffold.gbk"
file_list = (ref_ctg_file, scaff_gbk)
print "\t", g_name, "...",
# log
logstring = "".join(["\n", g_name])
run_ref.log(logstring)
# set outputs
mauve_dir = mauve_root+g_name+"/"
aln_segs_dir = segments_root+g_name+"/"
ensure_dir([mauve_dir, aln_segs_dir])
mauve_outfile = mauve_dir+g_name+"_"+ref_n+".mauve"
segfile = aln_segs_dir+g_name+"_"+ref_n+"_segs.txt"
# abort if the reference file is not found
try: open(ref_ctg_file, 'r')
except IOError:
msg = "ERROR: Reference file not found"
print msg
run_ref.log(msg)
raise
# abort if there is no scaffold construct
try: open(scaff_gbk, 'r')
except IOError:
msg = "WARNING: No scaffold construct to align"
print msg
run_ref.log(msg)
else:
# prep segments file
open(segfile, 'w').write('')
# purge any pre-existing sslist file
sslist_file = scaff_gbk+".sslist"
if os.path.isfile(sslist_file):
try: os.remove(sslist_file)
except Exception: raise
# do Mauve alignment
align_mauve(file_list, mauve_outfile, mauve_exec)
try:
# parse Mauve output (without initial clumping)
coords = mauver_load2_k0(mauve_outfile+".backbone", 0, mtype)
print len(coords), '->',
logstring = "".join(["\t", str(len(coords))])
run_ref.log(logstring)
# chop segments that are too long
chop_array = chop_rows(coords, max_size, chop_mode, mtype)
print len(chop_array), 'segments <', max_size, 'bp',
logstring = "".join(["\t", str(len(chop_array))])
run_ref.log(logstring)
# make detailed pairwise alignments of the segments
ref_rec = load_genbank(ref_ctg_file)
query_rec = load_genbank(scaff_gbk)
id = iter_align(chop_array, ref_rec, query_rec,
aln_segs_dir, segfile)
print "@", id, "% id. overall"
logstring = "".join(["\t", str(id)])
run_ref.log(logstring)
except IOError:
msg = "\nERROR: Mauve alignment failed"
run_ref.log(msg)
print msg
