def process_misassembled_contig(sorted_aligns, is_cyclic, aligned_lengths, region_misassemblies, ref_lens, ref_aligns,
ref_features, contig_seq, misassemblies_by_ref, istranslocations_by_ref, region_struct_variations,
ca_output):
misassembly_internal_overlap = 0
prev_align = sorted_aligns[0]
cur_aligned_length = prev_align.len2
is_misassembled = False
contig_is_printed = False
indels_info = IndelsInfo()
cnt_misassemblies = 0
misassemblies = []
misassembly_info = []
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
is_fake_translocation = is_fragmented_ref_fake_translocation(prev_align, next_align, ref_lens)
internal_overlap, overlap_msg = exclude_internal_overlaps(prev_align, next_align, i)
is_extensive_misassembly, aux_data = is_misassembly(prev_align, next_align, contig_seq, ref_lens,
is_cyclic, region_struct_variations, is_fake_translocation)
misassembly_type = ''
if is_extensive_misassembly: # it is not a Fake translocation, because is_extensive_misassembly is True
prev_ref, next_ref = get_ref_by_chromosome(prev_align.ref), get_ref_by_chromosome(next_align.ref)
if prev_align.ref != next_align.ref: # if chromosomes from different references
if qconfig.is_combined_ref and prev_ref != next_ref:
misassembly_type = 'interspecies translocation'
else:
misassembly_type = 'translocation'
elif abs(aux_data["inconsistency"]) > qconfig.extensive_misassembly_threshold:
misassembly_type = 'relocation'
else: #if strand1 != strand2:
misassembly_type = 'inversion'
if next_align.s1 > prev_align.e1:
start_in_ref, end_in_ref = prev_align.e1, next_align.s1
else:
start_in_ref, end_in_ref = next_align.s1, prev_align.e1
misassemblies.append([(prev_align, start_in_ref, misassembly_type, next_align.len2), (next_align, end_in_ref, misassembly_type)])
else:
misassemblies.append([])
misassembly_info.append((internal_overlap, overlap_msg, is_extensive_misassembly, aux_data, misassembly_type))
prev_align = next_align
is_potential_mge = None
if qconfig.large_genome:
is_potential_mge = detect_potential_mge(misassemblies)
prev_align = sorted_aligns[0]
contig_aligned_lengths = []
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
internal_overlap, overlap_msg, is_extensive_misassembly, aux_data, misassembly_type = misassembly_info[i]
if overlap_msg:
cur_aligned_length -= internal_overlap
ca_output.stdout_f.write(overlap_msg)
inconsistency = aux_data["inconsistency"]
distance_on_contig = aux_data["distance_on_contig"]
misassembly_internal_overlap += aux_data["misassembly_internal_overlap"]
cyclic_moment = aux_data["cyclic_moment"]
ca_output.icarus_out_f.write(prev_align.icarus_report_str() + '\n')
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s\n' % (i+1, str(prev_align)))
ref_aligns.setdefault(prev_align.ref, []).append(prev_align)
ca_output.coords_filtered_f.write(prev_align.coords_str() + '\n')
prev_ref, next_ref = get_ref_by_chromosome(prev_align.ref), get_ref_by_chromosome(next_align.ref)
if aux_data["is_sv"]:
ca_output.stdout_f.write('\t\t\t Not a misassembly (structural variation of the genome) between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly (structural variation of the genome)\n')
region_misassemblies.append(Misassembly.MATCHED_SV)
elif aux_data["is_scaffold_gap"]:
if abs(inconsistency) > qconfig.extensive_misassembly_threshold:
scaff_gap_type = ' (extensive)'
region_misassemblies.append(Misassembly.SCAFFOLD_GAP)
misassemblies_by_ref[prev_ref].append(Misassembly.SCAFFOLD_GAP)
ca_output.icarus_out_f.write('fake: scaffold gap size wrong estimation' + scaff_gap_type + '\n')
else:
scaff_gap_type = ' (local)'
region_misassemblies.append(Misassembly.LOCAL_SCAFFOLD_GAP)
misassemblies_by_ref[prev_ref].append(Misassembly.LOCAL_SCAFFOLD_GAP)
ca_output.icarus_out_f.write('fake: scaffold gap size wrong estimation' + scaff_gap_type + '\n')
ca_output.stdout_f.write('\t\t\t Scaffold gap between these two alignments, ')
ca_output.stdout_f.write('gap lengths difference (reference vs assembly) = ' + str(inconsistency) + scaff_gap_type + '\n')
elif is_extensive_misassembly and is_potential_mge and is_potential_mge[i]:
ca_output.stdout_f.write(
'\t\t\t Not a misassembly (possible transposable element) between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly (possible transposable element)\n')
region_misassemblies.append(Misassembly.POTENTIAL_MGE)
elif is_extensive_misassembly:
is_misassembled = True
cnt_misassemblies += 1
contig_aligned_lengths.append(cur_aligned_length)
cur_aligned_length = 0
if not contig_is_printed:
ca_output.misassembly_f.write(prev_align.contig + '\n')
contig_is_printed = True
ca_output.misassembly_f.write('Extensive misassembly (')
ca_output.stdout_f.write('\t\t\t Extensive misassembly (')
msg = ''
if misassembly_type == 'interspecies translocation':
misassembly_id = Misassembly.INTERSPECTRANSLOCATION
istranslocations_by_ref[prev_ref][next_ref] += 1
istranslocations_by_ref[next_ref][prev_ref] += 1
elif misassembly_type == 'translocation':
misassembly_id = Misassembly.TRANSLOCATION
elif misassembly_type == 'relocation':
misassembly_id = Misassembly.RELOCATION
msg = ', inconsistency = ' + str(inconsistency) + \
(' [linear representation of circular genome]' if cyclic_moment else '')
else: #if strand1 != strand2:
misassembly_id = Misassembly.INVERSION
region_misassemblies.append(misassembly_id)
misassemblies_by_ref[prev_ref].append(misassembly_id)
if misassembly_id == Misassembly.INTERSPECTRANSLOCATION: # special case
misassemblies_by_ref[next_ref].append(misassembly_id)
if is_gap_filled_ns(contig_seq, prev_align, next_align):
misassembly_type += ', scaffold gap is present'
region_misassemblies.append(misassembly_id + (Misassembly.SCF_INVERSION - Misassembly.INVERSION))
ca_output.stdout_f.write(misassembly_type + msg)
ca_output.misassembly_f.write(misassembly_type + msg)
ca_output.icarus_out_f.write(misassembly_type + msg)
ca_output.stdout_f.write(') between these two alignments\n')
ca_output.misassembly_f.write(') between %s %s and %s %s' % (prev_align.s2, prev_align.e2, next_align.s2, next_align.e2) + '\n')
ca_output.icarus_out_f.write('\n')
ref_features.setdefault(prev_align.ref, {})[prev_align.e1] = 'M'
ref_features.setdefault(next_align.ref, {})[next_align.e1] = 'M'
else:
reason_msg = "" + (" [linear representation of circular genome]" if cyclic_moment else "") + \
(" [fragmentation of reference genome]" if prev_align.ref != next_align.ref else "")
if inconsistency == 0 and cyclic_moment:
ca_output.stdout_f.write('\t\t\t Not a misassembly' + reason_msg + ' between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly' + reason_msg + '\n')
elif inconsistency == 0 and prev_align.ref != next_align.ref: # is_fragmented_ref_fake_translocation is True, because is_extensive_misassembly is False
ca_output.stdout_f.write('\t\t\t Not a misassembly' + reason_msg + ' between these two alignments\n')
region_misassemblies.append(Misassembly.FRAGMENTED)
misassemblies_by_ref[prev_ref].append(Misassembly.FRAGMENTED)
ca_output.icarus_out_f.write('fake: not a misassembly' + reason_msg + '\n')
elif abs(inconsistency) <= qconfig.MAX_INDEL_LENGTH and \
count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)[1] <= max(qconfig.min_alignment, qconfig.MAX_INDEL_LENGTH):
ns_number, not_ns_number = count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)
if inconsistency == 0:
ca_output.stdout_f.write(('\t\t\t Stretch of %d mismatches between these two alignments (number of Ns: %d)' %
(not_ns_number, ns_number)) + reason_msg + '\n')
indels_info.mismatches += not_ns_number
ca_output.icarus_out_f.write('indel: stretch of mismatches' + reason_msg + '\n')
else:
indel_length = abs(inconsistency)
indel_class = 'Indel (<= 5bp)' if indel_length <= qconfig.SHORT_INDEL_THRESHOLD else 'Indel (> 5bp)'
indel_type = 'insertion' if inconsistency < 0 else 'deletion'
mismatches = max(0, not_ns_number - indel_length)
ca_output.stdout_f.write(('\t\t\t %s between these two alignments: %s of length %d; %d mismatches (number of Ns: %d)')
% (indel_class, indel_type, indel_length, mismatches, ns_number) + reason_msg + '\n')
indels_info.indels_list.append(indel_length)
if indel_type == 'insertion':
indels_info.insertions += indel_length
else:
indels_info.deletions += indel_length
indels_info.mismatches += mismatches
ca_output.icarus_out_f.write('indel: ' + indel_class.lower() + reason_msg + '\n')
else:
if qconfig.strict_NA:
contig_aligned_lengths.append(cur_aligned_length)
cur_aligned_length = 0
if distance_on_contig < 0:
#There is an overlap between the two alignments, a local misassembly
ca_output.stdout_f.write('\t\t\t Overlap between these two alignments (local misassembly).')
elif distance_on_contig > 0:
#There is a small gap between the two alignments, a local misassembly
ca_output.stdout_f.write('\t\t\t Gap between these two alignments (local misassembly).')
elif inconsistency < 0:
ca_output.stdout_f.write('\t\t\t Overlap between these two alignments (local misassembly).')
else:
ca_output.stdout_f.write('\t\t\t Gap between these two alignments (local misassembly).')
ca_output.stdout_f.write(' Inconsistency = ' + str(inconsistency) + reason_msg + '\n')
ca_output.icarus_out_f.write('local misassembly' + reason_msg + '\n')
region_misassemblies.append(Misassembly.LOCAL)
misassemblies_by_ref[prev_ref].append(Misassembly.LOCAL)
prev_align = next_align
cur_aligned_length += prev_align.len2 - (-distance_on_contig if distance_on_contig < 0 else 0)
#Record the very last alignment
i = len(sorted_aligns) - 1
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s' % (i + 1, str(next_align)) + '\n')
ca_output.icarus_out_f.write(next_align.icarus_report_str() + '\n')
ref_aligns.setdefault(next_align.ref, []).append(next_align)
ca_output.coords_filtered_f.write(next_align.coords_str() + '\n')
contig_aligned_lengths.append(cur_aligned_length)
contig_aligned_length = sum(contig_aligned_lengths)
# if contig covers more than 95% of cyclic chromosome/plasmid consider it as cyclic and do not split the first and the last aligned blocks
if is_cyclic and len(contig_aligned_lengths) > 1 and sorted_aligns[-1].ref == sorted_aligns[0].ref and contig_aligned_length >= 0.95 * ref_lens[sorted_aligns[0].ref]:
is_extensive_misassembly, aux_data = is_misassembly(sorted_aligns[-1], sorted_aligns[0], contig_seq, ref_lens,
is_cyclic, is_cyclic_contig=True, region_struct_variations=region_struct_variations)
if not is_extensive_misassembly and not aux_data["is_scaffold_gap"] and not aux_data["is_sv"]:
inconsistency = abs(aux_data["inconsistency"])
if not qconfig.strict_NA or inconsistency <= qconfig.MAX_INDEL_LENGTH:
contig_aligned_lengths[0] += contig_aligned_lengths[-1]
contig_aligned_lengths = contig_aligned_lengths[:-1]
aligned_lengths.extend(contig_aligned_lengths)
assert contig_aligned_length <= len(contig_seq), "Internal QUAST bug: contig aligned length is greater than " \
"contig length (contig: %s, len: %d, aligned: %d)!" % \
(sorted_aligns[0].contig, contig_aligned_length, len(contig_seq))
return is_misassembled, misassembly_internal_overlap, indels_info, cnt_misassemblies, contig_aligned_length
def process_misassembled_contig(sorted_aligns, is_cyclic, aligned_lengths,
region_misassemblies, ref_lens, ref_aligns,
ref_features, contig_seq, misassemblies_by_ref,
istranslocations_by_ref,
region_struct_variations,
misassemblies_matched_sv, ca_output):
misassembly_internal_overlap = 0
prev_align = sorted_aligns[0]
cur_aligned_length = prev_align.len2
is_misassembled = False
contig_is_printed = False
indels_info = IndelsInfo()
contig_aligned_length = 0 # for internal debugging purposes
cnt_misassemblies = 0
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
is_fake_translocation = is_fragmented_ref_fake_translocation(
prev_align, next_align, ref_lens)
cur_aligned_length -= exclude_internal_overlaps(
prev_align, next_align, i, ca_output)
is_extensive_misassembly, aux_data = is_misassembly(
prev_align, next_align, contig_seq, ref_lens, is_cyclic,
region_struct_variations, is_fake_translocation)
inconsistency = aux_data["inconsistency"]
distance_on_contig = aux_data["distance_on_contig"]
misassembly_internal_overlap += aux_data[
"misassembly_internal_overlap"]
cyclic_moment = aux_data["cyclic_moment"]
ca_output.icarus_out_f.write(prev_align.icarus_report_str() + '\n')
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s\n' %
(i + 1, str(prev_align)))
ref_aligns.setdefault(prev_align.ref, []).append(prev_align)
ca_output.coords_filtered_f.write(str(prev_align) + '\n')
prev_ref, next_ref = get_ref_by_chromosome(
prev_align.ref), get_ref_by_chromosome(next_align.ref)
if aux_data["is_sv"]:
ca_output.stdout_f.write(
'\t\t\t Not a misassembly (structural variation of the genome) between these two alignments\n'
)
ca_output.icarus_out_f.write(
'fake: not a misassembly (structural variation of the genome)\n'
)
misassemblies_matched_sv += 1
elif aux_data["is_scaffold_gap"] and abs(
inconsistency) > qconfig.extensive_misassembly_threshold:
ca_output.stdout_f.write(
'\t\t\t Incorrectly estimated size of scaffold gap between these two alignments: '
)
ca_output.stdout_f.write('gap length difference = ' +
str(inconsistency) + '\n')
region_misassemblies.append(Misassembly.SCAFFOLD_GAP)
misassemblies_by_ref[prev_ref].append(Misassembly.SCAFFOLD_GAP)
ca_output.icarus_out_f.write(
'fake: scaffold gap size wrong estimation' + '\n')
elif is_extensive_misassembly:
is_misassembled = True
cnt_misassemblies += 1
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if not contig_is_printed:
ca_output.misassembly_f.write(prev_align.contig + '\n')
contig_is_printed = True
ca_output.misassembly_f.write('Extensive misassembly (')
ca_output.stdout_f.write('\t\t\t Extensive misassembly (')
if prev_align.ref != next_align.ref: # it is not a Fake translocation, because is_extensive_misassembly is True
if qconfig.is_combined_ref and prev_ref != next_ref: # if chromosomes from different references
region_misassemblies.append(
Misassembly.INTERSPECTRANSLOCATION)
istranslocations_by_ref[prev_ref][next_ref] += 1
istranslocations_by_ref[next_ref][prev_ref] += 1
misassemblies_by_ref[prev_ref].append(
Misassembly.INTERSPECTRANSLOCATION)
misassemblies_by_ref[next_ref].append(
Misassembly.INTERSPECTRANSLOCATION)
ca_output.stdout_f.write('interspecies translocation')
ca_output.misassembly_f.write('interspecies translocation')
ca_output.icarus_out_f.write('interspecies translocation')
else:
region_misassemblies.append(Misassembly.TRANSLOCATION)
misassemblies_by_ref[prev_ref].append(
Misassembly.TRANSLOCATION)
ca_output.stdout_f.write('translocation')
ca_output.misassembly_f.write('translocation')
ca_output.icarus_out_f.write('translocation')
elif abs(inconsistency) > qconfig.extensive_misassembly_threshold:
region_misassemblies.append(Misassembly.RELOCATION)
misassemblies_by_ref[prev_ref].append(Misassembly.RELOCATION)
msg = 'relocation, inconsistency = ' + str(inconsistency) + \
(' [linear representation of circular genome]' if cyclic_moment else '')
ca_output.stdout_f.write(msg)
ca_output.misassembly_f.write(msg)
ca_output.icarus_out_f.write(msg)
else: #if strand1 != strand2:
region_misassemblies.append(Misassembly.INVERSION)
misassemblies_by_ref[prev_ref].append(Misassembly.INVERSION)
ca_output.stdout_f.write('inversion')
ca_output.misassembly_f.write('inversion')
ca_output.icarus_out_f.write('inversion')
ca_output.stdout_f.write(') between these two alignments\n')
ca_output.misassembly_f.write(
') between %s %s and %s %s' %
(prev_align.s2, prev_align.e2, next_align.s2, next_align.e2) +
'\n')
ca_output.icarus_out_f.write('\n')
ref_features.setdefault(prev_align.ref, {})[prev_align.e1] = 'M'
ref_features.setdefault(next_align.ref, {})[next_align.e1] = 'M'
else:
reason_msg = "" + (" [linear representation of circular genome]" if cyclic_moment else "") + \
(" [fragmentation of reference genome]" if prev_align.ref != next_align.ref else "")
if inconsistency == 0 and cyclic_moment:
ca_output.stdout_f.write('\t\t\t Not a misassembly' +
reason_msg +
' between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly' +
reason_msg + '\n')
elif inconsistency == 0 and prev_align.ref != next_align.ref: # is_fragmented_ref_fake_translocation is True, because is_extensive_misassembly is False
ca_output.stdout_f.write('\t\t\t Not a misassembly' +
reason_msg +
' between these two alignments\n')
region_misassemblies.append(Misassembly.FRAGMENTED)
misassemblies_by_ref[prev_ref].append(Misassembly.FRAGMENTED)
ca_output.icarus_out_f.write('fake: not a misassembly' +
reason_msg + '\n')
elif abs(inconsistency) <= qconfig.MAX_INDEL_LENGTH and \
count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)[1] <= qconfig.MAX_INDEL_LENGTH:
ns_number, not_ns_number = count_ns_and_not_ns_between_aligns(
contig_seq, prev_align, next_align)
if inconsistency == 0:
ca_output.stdout_f.write((
'\t\t\t Stretch of %d mismatches between these two alignments (number of Ns: %d)'
% (not_ns_number, ns_number)) + reason_msg + '\n')
indels_info.mismatches += not_ns_number
ca_output.icarus_out_f.write(
'indel: stretch of mismatches' + reason_msg + '\n')
else:
indel_length = abs(inconsistency)
indel_class = 'Indel (<= 5bp)' if indel_length <= qconfig.SHORT_INDEL_THRESHOLD else 'Indel (> 5bp)'
indel_type = 'insertion' if inconsistency < 0 else 'deletion'
mismatches = max(0, not_ns_number - indel_length)
ca_output.stdout_f.write((
'\t\t\t %s between these two alignments: %s of length %d; %d mismatches (number of Ns: %d)'
) % (indel_class, indel_type, indel_length, mismatches,
ns_number) + reason_msg + '\n')
indels_info.indels_list.append(indel_length)
if indel_type == 'insertion':
indels_info.insertions += indel_length
else:
indels_info.deletions += indel_length
indels_info.mismatches += mismatches
ca_output.icarus_out_f.write('indel: ' +
indel_class.lower() +
reason_msg + '\n')
else:
if qconfig.strict_NA:
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if distance_on_contig < 0:
#There is an overlap between the two alignments, a local misassembly
ca_output.stdout_f.write(
'\t\t\t Overlap between these two alignments (local misassembly).'
)
elif distance_on_contig > 0:
#There is a small gap between the two alignments, a local misassembly
ca_output.stdout_f.write(
'\t\t\t Gap between these two alignments (local misassembly).'
)
elif inconsistency < 0:
ca_output.stdout_f.write(
'\t\t\t Overlap between these two alignments (local misassembly).'
)
else:
ca_output.stdout_f.write(
'\t\t\t Gap between these two alignments (local misassembly).'
)
ca_output.stdout_f.write(' Inconsistency = ' +
str(inconsistency) + reason_msg +
'\n')
ca_output.icarus_out_f.write('local misassembly' + reason_msg +
'\n')
region_misassemblies.append(Misassembly.LOCAL)
misassemblies_by_ref[prev_ref].append(Misassembly.LOCAL)
prev_align = next_align
cur_aligned_length += prev_align.len2 - (-distance_on_contig if
distance_on_contig < 0 else 0)
#Record the very last alignment
i = len(sorted_aligns) - 1
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s' %
(i + 1, str(next_align)) + '\n')
ca_output.icarus_out_f.write(next_align.icarus_report_str() + '\n')
ref_aligns.setdefault(next_align.ref, []).append(next_align)
ca_output.coords_filtered_f.write(str(next_align) + '\n')
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
assert contig_aligned_length <= len(contig_seq), "Internal QUAST bug: contig aligned length is greater than " \
"contig length (contig: %s, len: %d, aligned: %d)!" % \
(sorted_aligns[0].contig, contig_aligned_length, len(contig_seq))
return is_misassembled, misassembly_internal_overlap, indels_info, misassemblies_matched_sv, cnt_misassemblies, contig_aligned_length
def process_misassembled_contig(sorted_aligns, cyclic, aligned_lengths, region_misassemblies, ref_lens, ref_aligns,
ref_features, contig_seq, references_misassemblies, region_struct_variations,
misassemblies_matched_sv, ca_output, is_ambiguous=False):
misassembly_internal_overlap = 0
prev_align = sorted_aligns[0]
cur_aligned_length = prev_align[5]
is_misassembled = False
contig_is_printed = False
indels_info = IndelsInfo()
contig_aligned_length = 0 # for internal debugging purposes
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
cur_aligned_length -= exclude_internal_overlaps(prev_align, next_align, i, ca_output)
is_extensive_misassembly, aux_data = is_misassembly(prev_align, next_align, contig_seq, ref_lens,
cyclic, region_struct_variations)
inconsistency = aux_data["inconsistency"]
distance_on_contig = aux_data["distance_on_contig"]
misassembly_internal_overlap += aux_data["misassembly_internal_overlap"]
cyclic_moment = aux_data["cyclic_moment"]
# print >> ca_output.icarus_out_f, prev_align.icarus_report_str()
print >> ca_output.icarus_out_f, icarus_report_str(prev_align)
print >> ca_output.stdout_f, '\t\t\tReal Alignment %d: %s' % (i+1, str(prev_align))
ref_aligns.setdefault(prev_align[7], []).append(prev_align)
print >> ca_output.coords_filtered_f, str(prev_align)
if aux_data["is_sv"]:
print >> ca_output.stdout_f, '\t\t\t Not a misassembly (structural variation of the genome) between these two alignments'
print >> ca_output.icarus_out_f, 'fake: not a misassembly (structural variation of the genome)'
misassemblies_matched_sv += 1
elif aux_data["is_scaffold_gap"]:
print >> ca_output.stdout_f, '\t\t\t Incorrectly estimated size of scaffold gap between these two alignments:',
print >> ca_output.stdout_f, 'gap length difference =', inconsistency
region_misassemblies.append(Misassembly.SCAFFOLD_GAP)
print >> ca_output.icarus_out_f, 'fake: scaffold gap size wrong estimation'
elif is_extensive_misassembly:
is_misassembled = True
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if not contig_is_printed:
print >> ca_output.misassembly_f, prev_align[8]
contig_is_printed = True
print >> ca_output.misassembly_f, 'Extensive misassembly (',
print >> ca_output.stdout_f, '\t\t\t Extensive misassembly (',
if prev_align[7] != next_align[7]: # it is not a Fake translocation, because is_extensive_misassembly is True
if qconfig.is_combined_ref and \
not is_same_reference(prev_align[7], next_align[7]): # if chromosomes from different references
region_misassemblies.append(Misassembly.INTERSPECTRANSLOCATION)
ref1, ref2 = get_ref_by_chromosome(prev_align[7]), get_ref_by_chromosome(next_align[7])
references_misassemblies[ref1][ref2] += 1
references_misassemblies[ref2][ref1] += 1
print >> ca_output.stdout_f, 'interspecies translocation',
print >> ca_output.misassembly_f, 'interspecies translocation',
print >> ca_output.icarus_out_f, 'interspecies translocation'
else:
region_misassemblies.append(Misassembly.TRANSLOCATION)
print >> ca_output.stdout_f, 'translocation',
print >> ca_output.misassembly_f, 'translocation',
print >> ca_output.icarus_out_f, 'translocation'
elif abs(inconsistency) > qconfig.extensive_misassembly_threshold:
region_misassemblies.append(Misassembly.RELOCATION)
print >> ca_output.stdout_f, 'relocation, inconsistency =', inconsistency,
print >> ca_output.misassembly_f, 'relocation, inconsistency =', inconsistency,
print >> ca_output.icarus_out_f, 'relocation, inconsistency =', inconsistency
else: #if strand1 != strand2:
region_misassemblies.append(Misassembly.INVERSION)
print >> ca_output.stdout_f, 'inversion',
print >> ca_output.misassembly_f, 'inversion',
print >> ca_output.icarus_out_f, 'inversion'
print >> ca_output.stdout_f, ') between these two alignments'
print >> ca_output.misassembly_f, ') between %s %s and %s %s' % (prev_align[2], prev_align[3],
next_align[2], next_align[3])
ref_features.setdefault(prev_align[7], {})[prev_align[1]] = 'M'
ref_features.setdefault(next_align[7], {})[next_align[1]] = 'M'
else:
reason_msg = "" + (" (linear representation of circular genome)" if cyclic_moment else "") + \
(" (fragmentation of reference genome)" if prev_align[7] != next_align[7] else "")
if inconsistency == 0 and cyclic_moment:
print >> ca_output.stdout_f, '\t\t\t Not a misassembly' + reason_msg + ' between these two alignments'
print >> ca_output.icarus_out_f, 'fake: not a misassembly' + reason_msg
elif inconsistency == 0 and prev_align[7] != next_align[7]: # is_fragmented_ref_fake_translocation is True, because is_extensive_misassembly is False
print >> ca_output.stdout_f, '\t\t\t Not a misassembly' + reason_msg + ' between these two alignments'
region_misassemblies.append(Misassembly.FRAGMENTED)
print >> ca_output.icarus_out_f, 'fake: not a misassembly' + reason_msg
elif abs(inconsistency) <= qconfig.MAX_INDEL_LENGTH and \
count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)[1] <= qconfig.MAX_INDEL_LENGTH:
ns_number, not_ns_number = count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)
if inconsistency == 0:
print >> ca_output.stdout_f, ('\t\t\t Short stretch of %d mismatches and %d Ns between these two alignments' % (not_ns_number, ns_number)) + reason_msg
indels_info.mismatches += not_ns_number
print >> ca_output.icarus_out_f, 'indel: stretch of mismatches and Ns' + reason_msg
else:
indel_length = abs(inconsistency)
indel_class = 'Indel (<= 5bp)' if indel_length <= qconfig.SHORT_INDEL_THRESHOLD else 'Indel (> 5bp)'
indel_type = 'insertion' if inconsistency < 0 else 'deletion'
mismatches = max(0, not_ns_number - indel_length)
print >> ca_output.stdout_f, ('\t\t\t %s between these two alignments: %s of length %d; %d mismatches'
% (indel_class, indel_type, indel_length, mismatches)) + reason_msg
indels_info.indels_list.append(indel_length)
if indel_type == 'insertion':
indels_info.insertions += indel_length
else:
indels_info.deletions += indel_length
indels_info.mismatches += mismatches
print >> ca_output.icarus_out_f, 'indel: ' + indel_class.lower() + reason_msg
else:
if qconfig.strict_NA:
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if inconsistency < 0:
#There is an overlap between the two alignments, a local misassembly
print >> ca_output.stdout_f, '\t\t\t Overlap between these two alignments (local misassembly).',
else:
#There is a small gap between the two alignments, a local misassembly
print >> ca_output.stdout_f, '\t\t\t Gap between these two alignments (local misassembly).',
#print >> plantafile_out, 'Distance on contig =', distance_on_contig, ', distance on reference =', distance_on_reference
print >> ca_output.stdout_f, 'Inconsistency = ' + str(inconsistency) + reason_msg
print >> ca_output.icarus_out_f, 'local misassembly'
region_misassemblies.append(Misassembly.LOCAL)
prev_align = next_align
cur_aligned_length += prev_align[5] - (-distance_on_contig if distance_on_contig < 0 else 0)
#Record the very last alignment
i = len(sorted_aligns) - 1
print >> ca_output.stdout_f, '\t\t\tReal Alignment %d: %s' % (i + 1, str(next_align))
# print >> ca_output.icarus_out_f, next_align.icarus_report_str()
print >> ca_output.icarus_out_f, icarus_report_str(next_align)
ref_aligns.setdefault(next_align[7], []).append(next_align)
print >> ca_output.coords_filtered_f, str(next_align)
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
assert contig_aligned_length <= len(contig_seq), "Internal QUAST bug: contig aligned length is greater than " \
"contig length (contig: %s, len: %d, aligned: %d)!" % \
(sorted_aligns[0][8], contig_aligned_length, len(contig_seq))
return is_misassembled, misassembly_internal_overlap, references_misassemblies, indels_info, misassemblies_matched_sv
def process_misassembled_contig(sorted_aligns, cyclic, aligned_lengths, region_misassemblies, ref_lens, ref_aligns,
ref_features, contig_seq, references_misassemblies, region_struct_variations,
misassemblies_matched_sv, ca_output, is_ambiguous=False):
misassembly_internal_overlap = 0
prev_align = sorted_aligns[0]
cur_aligned_length = prev_align.len2
is_misassembled = False
contig_is_printed = False
indels_info = IndelsInfo()
contig_aligned_length = 0 # for internal debugging purposes
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
is_fake_translocation = is_fragmented_ref_fake_translocation(prev_align, next_align, ref_lens)
cur_aligned_length -= exclude_internal_overlaps(prev_align, next_align, i, ca_output)
is_extensive_misassembly, aux_data = is_misassembly(prev_align, next_align, contig_seq, ref_lens,
cyclic, region_struct_variations, is_fake_translocation)
inconsistency = aux_data["inconsistency"]
distance_on_contig = aux_data["distance_on_contig"]
misassembly_internal_overlap += aux_data["misassembly_internal_overlap"]
cyclic_moment = aux_data["cyclic_moment"]
ca_output.icarus_out_f.write(prev_align.icarus_report_str() + '\n')
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s\n' % (i+1, str(prev_align)))
ref_aligns.setdefault(prev_align.ref, []).append(prev_align)
ca_output.coords_filtered_f.write(str(prev_align) + '\n')
if aux_data["is_sv"]:
ca_output.stdout_f.write('\t\t\t Not a misassembly (structural variation of the genome) between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly (structural variation of the genome)\n')
misassemblies_matched_sv += 1
elif aux_data["is_scaffold_gap"] and abs(inconsistency) > qconfig.extensive_misassembly_threshold:
ca_output.stdout_f.write('\t\t\t Incorrectly estimated size of scaffold gap between these two alignments: ')
ca_output.stdout_f.write('gap length difference = ' + str(inconsistency) + '\n')
region_misassemblies.append(Misassembly.SCAFFOLD_GAP)
ca_output.icarus_out_f.write('fake: scaffold gap size wrong estimation' + '\n')
elif is_extensive_misassembly:
is_misassembled = True
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if not contig_is_printed:
ca_output.misassembly_f.write(prev_align.contig + '\n')
contig_is_printed = True
ca_output.misassembly_f.write('Extensive misassembly (')
ca_output.stdout_f.write('\t\t\t Extensive misassembly (')
if prev_align.ref != next_align.ref: # it is not a Fake translocation, because is_extensive_misassembly is True
if qconfig.is_combined_ref and \
not is_same_reference(prev_align.ref, next_align.ref): # if chromosomes from different references
region_misassemblies.append(Misassembly.INTERSPECTRANSLOCATION)
ref1, ref2 = get_ref_by_chromosome(prev_align.ref), get_ref_by_chromosome(next_align.ref)
references_misassemblies[ref1][ref2] += 1
references_misassemblies[ref2][ref1] += 1
ca_output.stdout_f.write('interspecies translocation')
ca_output.misassembly_f.write('interspecies translocation')
ca_output.icarus_out_f.write('interspecies translocation')
else:
region_misassemblies.append(Misassembly.TRANSLOCATION)
ca_output.stdout_f.write('translocation')
ca_output.misassembly_f.write('translocation')
ca_output.icarus_out_f.write('translocation')
elif abs(inconsistency) > qconfig.extensive_misassembly_threshold:
region_misassemblies.append(Misassembly.RELOCATION)
msg = 'relocation, inconsistency = ' + str(inconsistency) + \
(' [linear representation of circular genome]' if cyclic_moment else '')
ca_output.stdout_f.write(msg)
ca_output.misassembly_f.write(msg)
ca_output.icarus_out_f.write(msg)
else: #if strand1 != strand2:
region_misassemblies.append(Misassembly.INVERSION)
ca_output.stdout_f.write('inversion')
ca_output.misassembly_f.write('inversion')
ca_output.icarus_out_f.write('inversion')
ca_output.stdout_f.write(') between these two alignments\n')
ca_output.misassembly_f.write(') between %s %s and %s %s' % (prev_align.s2, prev_align.e2, next_align.s2, next_align.e2) + '\n')
ca_output.icarus_out_f.write('\n')
ref_features.setdefault(prev_align.ref, {})[prev_align.e1] = 'M'
ref_features.setdefault(next_align.ref, {})[next_align.e1] = 'M'
else:
reason_msg = "" + (" [linear representation of circular genome]" if cyclic_moment else "") + \
(" [fragmentation of reference genome]" if prev_align.ref != next_align.ref else "")
if inconsistency == 0 and cyclic_moment:
ca_output.stdout_f.write('\t\t\t Not a misassembly' + reason_msg + ' between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly' + reason_msg + '\n')
elif inconsistency == 0 and prev_align.ref != next_align.ref: # is_fragmented_ref_fake_translocation is True, because is_extensive_misassembly is False
ca_output.stdout_f.write('\t\t\t Not a misassembly' + reason_msg + ' between these two alignments\n')
region_misassemblies.append(Misassembly.FRAGMENTED)
ca_output.icarus_out_f.write('fake: not a misassembly' + reason_msg + '\n')
elif abs(inconsistency) <= qconfig.MAX_INDEL_LENGTH and \
count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)[1] <= qconfig.MAX_INDEL_LENGTH:
ns_number, not_ns_number = count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)
if inconsistency == 0:
ca_output.stdout_f.write(('\t\t\t Stretch of %d mismatches between these two alignments (number of Ns: %d)' %
(not_ns_number, ns_number)) + reason_msg + '\n')
indels_info.mismatches += not_ns_number
ca_output.icarus_out_f.write('indel: stretch of mismatches' + reason_msg + '\n')
else:
indel_length = abs(inconsistency)
indel_class = 'Indel (<= 5bp)' if indel_length <= qconfig.SHORT_INDEL_THRESHOLD else 'Indel (> 5bp)'
indel_type = 'insertion' if inconsistency < 0 else 'deletion'
mismatches = max(0, not_ns_number - indel_length)
ca_output.stdout_f.write(('\t\t\t %s between these two alignments: %s of length %d; %d mismatches (number of Ns: %d)')
% (indel_class, indel_type, indel_length, mismatches, ns_number) + reason_msg + '\n')
indels_info.indels_list.append(indel_length)
if indel_type == 'insertion':
indels_info.insertions += indel_length
else:
indels_info.deletions += indel_length
indels_info.mismatches += mismatches
ca_output.icarus_out_f.write('indel: ' + indel_class.lower() + reason_msg + '\n')
else:
if qconfig.strict_NA:
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if distance_on_contig < 0:
#There is an overlap between the two alignments, a local misassembly
ca_output.stdout_f.write('\t\t\t Overlap between these two alignments (local misassembly).')
elif distance_on_contig > 0:
#There is a small gap between the two alignments, a local misassembly
ca_output.stdout_f.write('\t\t\t Gap between these two alignments (local misassembly).')
elif inconsistency < 0:
ca_output.stdout_f.write('\t\t\t Overlap between these two alignments (local misassembly).')
else:
ca_output.stdout_f.write('\t\t\t Gap between these two alignments (local misassembly).')
ca_output.stdout_f.write(' Inconsistency = ' + str(inconsistency) + reason_msg + '\n')
ca_output.icarus_out_f.write('local misassembly' + reason_msg + '\n')
region_misassemblies.append(Misassembly.LOCAL)
prev_align = next_align
cur_aligned_length += prev_align.len2 - (-distance_on_contig if distance_on_contig < 0 else 0)
#Record the very last alignment
i = len(sorted_aligns) - 1
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s' % (i + 1, str(next_align)) + '\n')
ca_output.icarus_out_f.write(next_align.icarus_report_str() + '\n')
ref_aligns.setdefault(next_align.ref, []).append(next_align)
ca_output.coords_filtered_f.write(str(next_align) + '\n')
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
assert contig_aligned_length <= len(contig_seq), "Internal QUAST bug: contig aligned length is greater than " \
"contig length (contig: %s, len: %d, aligned: %d)!" % \
(sorted_aligns[0].contig, contig_aligned_length, len(contig_seq))
return is_misassembled, misassembly_internal_overlap, references_misassemblies, indels_info, misassemblies_matched_sv
def process_misassembled_contig(sorted_aligns, is_cyclic, aligned_lengths, region_misassemblies, ref_lens, ref_aligns,
ref_features, contig_seq, misassemblies_by_ref, istranslocations_by_ref, region_struct_variations,
ca_output):
misassembly_internal_overlap = 0
prev_align = sorted_aligns[0]
cur_aligned_length = prev_align.len2
is_misassembled = False
contig_is_printed = False
indels_info = IndelsInfo()
contig_aligned_length = 0 # for internal debugging purposes
cnt_misassemblies = 0
misassemblies = []
misassembly_info = []
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
is_fake_translocation = is_fragmented_ref_fake_translocation(prev_align, next_align, ref_lens)
internal_overlap, overlap_msg = exclude_internal_overlaps(prev_align, next_align, i)
is_extensive_misassembly, aux_data = is_misassembly(prev_align, next_align, contig_seq, ref_lens,
is_cyclic, region_struct_variations, is_fake_translocation)
misassembly_type = ''
if is_extensive_misassembly: # it is not a Fake translocation, because is_extensive_misassembly is True
prev_ref, next_ref = get_ref_by_chromosome(prev_align.ref), get_ref_by_chromosome(next_align.ref)
if prev_align.ref != next_align.ref: # if chromosomes from different references
if qconfig.is_combined_ref and prev_ref != next_ref:
misassembly_type = 'interspecies translocation'
else:
misassembly_type = 'translocation'
elif abs(aux_data["inconsistency"]) > qconfig.extensive_misassembly_threshold:
misassembly_type = 'relocation'
else: #if strand1 != strand2:
misassembly_type = 'inversion'
if next_align.s1 > prev_align.e1:
start_in_ref, end_in_ref = prev_align.e1, next_align.s1
else:
start_in_ref, end_in_ref = next_align.s1, prev_align.e1
misassemblies.append([(prev_align, start_in_ref, misassembly_type, next_align.len2), (next_align, end_in_ref, misassembly_type)])
else:
misassemblies.append([])
misassembly_info.append((internal_overlap, overlap_msg, is_extensive_misassembly, aux_data, misassembly_type))
prev_align = next_align
is_potential_mge = None
if qconfig.large_genome:
is_potential_mge = detect_potential_mge(misassemblies)
prev_align = sorted_aligns[0]
for i in range(len(sorted_aligns) - 1):
next_align = sorted_aligns[i + 1]
internal_overlap, overlap_msg, is_extensive_misassembly, aux_data, misassembly_type = misassembly_info[i]
if overlap_msg:
cur_aligned_length -= internal_overlap
ca_output.stdout_f.write(overlap_msg)
inconsistency = aux_data["inconsistency"]
distance_on_contig = aux_data["distance_on_contig"]
misassembly_internal_overlap += aux_data["misassembly_internal_overlap"]
cyclic_moment = aux_data["cyclic_moment"]
ca_output.icarus_out_f.write(prev_align.icarus_report_str() + '\n')
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s\n' % (i+1, str(prev_align)))
ref_aligns.setdefault(prev_align.ref, []).append(prev_align)
ca_output.coords_filtered_f.write(prev_align.coords_str() + '\n')
prev_ref, next_ref = get_ref_by_chromosome(prev_align.ref), get_ref_by_chromosome(next_align.ref)
if aux_data["is_sv"]:
ca_output.stdout_f.write('\t\t\t Not a misassembly (structural variation of the genome) between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly (structural variation of the genome)\n')
region_misassemblies.append(Misassembly.MATCHED_SV)
elif aux_data["is_scaffold_gap"]:
if abs(inconsistency) > qconfig.extensive_misassembly_threshold:
scaff_gap_type = ' (extensive)'
region_misassemblies.append(Misassembly.SCAFFOLD_GAP)
misassemblies_by_ref[prev_ref].append(Misassembly.SCAFFOLD_GAP)
ca_output.icarus_out_f.write('fake: scaffold gap size wrong estimation' + scaff_gap_type + '\n')
else:
scaff_gap_type = ' (local)'
region_misassemblies.append(Misassembly.LOCAL_SCAFFOLD_GAP)
misassemblies_by_ref[prev_ref].append(Misassembly.LOCAL_SCAFFOLD_GAP)
ca_output.icarus_out_f.write('fake: scaffold gap size wrong estimation' + scaff_gap_type + '\n')
ca_output.stdout_f.write('\t\t\t Scaffold gap between these two alignments, ')
ca_output.stdout_f.write('gap lengths difference (reference vs assembly) = ' + str(inconsistency) + scaff_gap_type + '\n')
elif is_extensive_misassembly and is_potential_mge and is_potential_mge[i]:
ca_output.stdout_f.write(
'\t\t\t Not a misassembly (possible transposable element) between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly (possible transposable element)\n')
region_misassemblies.append(Misassembly.POTENTIAL_MGE)
elif is_extensive_misassembly:
is_misassembled = True
cnt_misassemblies += 1
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if not contig_is_printed:
ca_output.misassembly_f.write(prev_align.contig + '\n')
contig_is_printed = True
ca_output.misassembly_f.write('Extensive misassembly (')
ca_output.stdout_f.write('\t\t\t Extensive misassembly (')
msg = ''
if misassembly_type == 'interspecies translocation':
misassembly_id = Misassembly.INTERSPECTRANSLOCATION
istranslocations_by_ref[prev_ref][next_ref] += 1
istranslocations_by_ref[next_ref][prev_ref] += 1
elif misassembly_type == 'translocation':
misassembly_id = Misassembly.TRANSLOCATION
elif misassembly_type == 'relocation':
misassembly_id = Misassembly.RELOCATION
msg = ', inconsistency = ' + str(inconsistency) + \
(' [linear representation of circular genome]' if cyclic_moment else '')
else: #if strand1 != strand2:
misassembly_id = Misassembly.INVERSION
region_misassemblies.append(misassembly_id)
misassemblies_by_ref[prev_ref].append(misassembly_id)
if misassembly_id == Misassembly.INTERSPECTRANSLOCATION: # special case
misassemblies_by_ref[next_ref].append(misassembly_id)
if is_gap_filled_ns(contig_seq, prev_align, next_align):
misassembly_type += ', scaffold gap is present'
region_misassemblies.append(misassembly_id + (Misassembly.SCF_INVERSION - Misassembly.INVERSION))
ca_output.stdout_f.write(misassembly_type + msg)
ca_output.misassembly_f.write(misassembly_type + msg)
ca_output.icarus_out_f.write(misassembly_type + msg)
ca_output.stdout_f.write(') between these two alignments\n')
ca_output.misassembly_f.write(') between %s %s and %s %s' % (prev_align.s2, prev_align.e2, next_align.s2, next_align.e2) + '\n')
ca_output.icarus_out_f.write('\n')
ref_features.setdefault(prev_align.ref, {})[prev_align.e1] = 'M'
ref_features.setdefault(next_align.ref, {})[next_align.e1] = 'M'
else:
reason_msg = "" + (" [linear representation of circular genome]" if cyclic_moment else "") + \
(" [fragmentation of reference genome]" if prev_align.ref != next_align.ref else "")
if inconsistency == 0 and cyclic_moment:
ca_output.stdout_f.write('\t\t\t Not a misassembly' + reason_msg + ' between these two alignments\n')
ca_output.icarus_out_f.write('fake: not a misassembly' + reason_msg + '\n')
elif inconsistency == 0 and prev_align.ref != next_align.ref: # is_fragmented_ref_fake_translocation is True, because is_extensive_misassembly is False
ca_output.stdout_f.write('\t\t\t Not a misassembly' + reason_msg + ' between these two alignments\n')
region_misassemblies.append(Misassembly.FRAGMENTED)
misassemblies_by_ref[prev_ref].append(Misassembly.FRAGMENTED)
ca_output.icarus_out_f.write('fake: not a misassembly' + reason_msg + '\n')
elif abs(inconsistency) <= qconfig.MAX_INDEL_LENGTH and \
count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)[1] <= max(qconfig.min_alignment, qconfig.MAX_INDEL_LENGTH):
ns_number, not_ns_number = count_ns_and_not_ns_between_aligns(contig_seq, prev_align, next_align)
if inconsistency == 0:
ca_output.stdout_f.write(('\t\t\t Stretch of %d mismatches between these two alignments (number of Ns: %d)' %
(not_ns_number, ns_number)) + reason_msg + '\n')
indels_info.mismatches += not_ns_number
ca_output.icarus_out_f.write('indel: stretch of mismatches' + reason_msg + '\n')
else:
indel_length = abs(inconsistency)
indel_class = 'Indel (<= 5bp)' if indel_length <= qconfig.SHORT_INDEL_THRESHOLD else 'Indel (> 5bp)'
indel_type = 'insertion' if inconsistency < 0 else 'deletion'
mismatches = max(0, not_ns_number - indel_length)
ca_output.stdout_f.write(('\t\t\t %s between these two alignments: %s of length %d; %d mismatches (number of Ns: %d)')
% (indel_class, indel_type, indel_length, mismatches, ns_number) + reason_msg + '\n')
indels_info.indels_list.append(indel_length)
if indel_type == 'insertion':
indels_info.insertions += indel_length
else:
indels_info.deletions += indel_length
indels_info.mismatches += mismatches
ca_output.icarus_out_f.write('indel: ' + indel_class.lower() + reason_msg + '\n')
else:
if qconfig.strict_NA:
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
cur_aligned_length = 0
if distance_on_contig < 0:
#There is an overlap between the two alignments, a local misassembly
ca_output.stdout_f.write('\t\t\t Overlap between these two alignments (local misassembly).')
elif distance_on_contig > 0:
#There is a small gap between the two alignments, a local misassembly
ca_output.stdout_f.write('\t\t\t Gap between these two alignments (local misassembly).')
elif inconsistency < 0:
ca_output.stdout_f.write('\t\t\t Overlap between these two alignments (local misassembly).')
else:
ca_output.stdout_f.write('\t\t\t Gap between these two alignments (local misassembly).')
ca_output.stdout_f.write(' Inconsistency = ' + str(inconsistency) + reason_msg + '\n')
ca_output.icarus_out_f.write('local misassembly' + reason_msg + '\n')
region_misassemblies.append(Misassembly.LOCAL)
misassemblies_by_ref[prev_ref].append(Misassembly.LOCAL)
prev_align = next_align
cur_aligned_length += prev_align.len2 - (-distance_on_contig if distance_on_contig < 0 else 0)
#Record the very last alignment
i = len(sorted_aligns) - 1
ca_output.stdout_f.write('\t\t\tReal Alignment %d: %s' % (i + 1, str(next_align)) + '\n')
ca_output.icarus_out_f.write(next_align.icarus_report_str() + '\n')
ref_aligns.setdefault(next_align.ref, []).append(next_align)
ca_output.coords_filtered_f.write(next_align.coords_str() + '\n')
aligned_lengths.append(cur_aligned_length)
contig_aligned_length += cur_aligned_length
assert contig_aligned_length <= len(contig_seq), "Internal QUAST bug: contig aligned length is greater than " \
"contig length (contig: %s, len: %d, aligned: %d)!" % \
(sorted_aligns[0].contig, contig_aligned_length, len(contig_seq))
return is_misassembled, misassembly_internal_overlap, indels_info, cnt_misassemblies, contig_aligned_length
