def add_contig_structure_data(data_str, structure, ref_contigs, chr_full_names, contig_names_by_refs,
used_chromosomes, links_to_chromosomes, chr_names_by_id):
for el in structure:
if isinstance(el, Alignment):
if el.ref_name in ref_contigs:
num_chr = ref_contigs.index(el.ref_name)
# corr_len = sum(chr_lengths[:num_chr+1])
else:
# corr_len = -int(el.end)
if el.ref_name not in used_chromosomes:
used_chromosomes.append(el.ref_name)
if contig_names_by_refs:
other_ref_name = contig_names_by_refs[el.ref_name]
links_to_chromosomes.append('links_to_chromosomes["' + el.ref_name + '"] = "' +
get_html_name(other_ref_name, chr_full_names) + '";')
corr_el_start = el.start
corr_el_end = el.end
data_str.append('{corr_start: ' + str(corr_el_start) + ',corr_end: ' +
str(corr_el_end) + ',start:' + str(el.unshifted_start) + ',end:' + str(el.unshifted_end) +
',start_in_contig:' + str(el.start_in_contig) + ',end_in_contig:' +
str(el.end_in_contig) + ',IDY:' + el.idy + ',chr: "' + chr_names_by_id[el.ref_name] + '"},')
elif type(el) == str:
ms_description, ms_type = parse_misassembly_info(el)
data_str.append('{contig_type: "M", mstype: "' + ms_type + '", msg: "' + ms_description + '"},')
return data_str
def prepare_alignment_data_for_one_ref(chr, chr_full_names, chr_names_by_id, ref_contigs, data_str, chr_to_aligned_blocks,
structures_by_labels, contigs_by_assemblies, ambiguity_alignments_by_labels=None,
contig_names_by_refs=None, output_dir_path=None,
cov_data_str=None, physical_cov_data_str=None, gc_data_str=None):
html_name = get_html_name(chr, chr_full_names)
alignment_viewer_fpath = join(output_dir_path, html_name + '.html')
additional_assemblies_data = ''
data_str.append('var links_to_chromosomes;')
links_to_chromosomes = []
used_chromosomes = []
if contig_names_by_refs:
data_str.append('links_to_chromosomes = {};')
num_misassemblies = 0
aligned_assemblies = set()
is_one_html = len(chr_full_names) == 1
data_str.append('var oneHtml = ' + str(is_one_html).lower() + ';')
# adding assembly data
data_str.append('var contig_data = {};')
data_str.append('contig_data["' + chr + '"] = {};')
assemblies_len = defaultdict(int)
assemblies_contigs = defaultdict(set)
ms_types = dict()
for assembly in chr_to_aligned_blocks.keys():
data_str.append('contig_data["' + chr + '"]["' + assembly + '"] = [ ')
ms_types[assembly] = defaultdict(int)
contigs = dict((contig.name, contig) for contig in contigs_by_assemblies[assembly])
for num_contig, ref_contig in enumerate(ref_contigs):
if ref_contig in chr_to_aligned_blocks[assembly]:
overlapped_contigs = defaultdict(list)
alignments = sorted(chr_to_aligned_blocks[assembly][ref_contig], key=lambda x: x.start)
prev_end = 0
prev_alignments = []
for alignment in alignments:
if prev_end > alignment.start:
for prev_align in prev_alignments:
if alignment.name != prev_align.name and prev_align.end - alignment.start > 100:
overlapped_contigs[prev_align].append('{contig:"' + alignment.name + '",corr_start: ' + str(alignment.start) +
',corr_end: ' + str(alignment.end) + ',start:' + str(alignment.unshifted_start) +
',end:' + str(alignment.unshifted_end) + ',start_in_contig:' + str(alignment.start_in_contig) +
',end_in_contig:' + str(alignment.end_in_contig) + ',chr: "' + chr_names_by_id[alignment.ref_name] + '"}')
overlapped_contigs[alignment].append('{contig:"' + prev_align.name + '",corr_start: ' + str(prev_align.start) +
',corr_end: ' + str(prev_align.end) + ',start:' + str(prev_align.unshifted_start) +
',end:' + str(prev_align.unshifted_end) + ',start_in_contig:' + str(prev_align.start_in_contig) +
',end_in_contig:' + str(prev_align.end_in_contig) + ',chr: "' + chr_names_by_id[prev_align.ref_name] + '"}')
prev_alignments.append(alignment)
else:
prev_alignments = [alignment]
prev_end = max(prev_end, alignment.end)
for alignment in alignments:
assemblies_len[assembly] += abs(alignment.end_in_contig - alignment.start_in_contig) + 1
assemblies_contigs[assembly].add(alignment.name)
contig_structure = structures_by_labels[alignment.label]
contig_more_unaligned = False
if alignment.misassembled:
num_misassemblies += 1
misassemblies, misassembled_ends = get_misassembly_for_alignment(contig_structure[alignment.name], alignment)
for misassembly in misassemblies:
if misassembly:
ms_types[assembly][misassembly] += 1
alignment.misassemblies = ';'.join(misassemblies)
if 'unknown' in alignment.misassemblies:
alignment.misassemblies = 'unknown'
misassembled_ends = ''
contig_more_unaligned = True
else:
if contigs[alignment.name].contig_type == 'correct_unaligned':
contig_more_unaligned = True
misassembled_ends = ''
genes = []
start_in_contig, end_in_contig = min(alignment.start_in_contig, alignment.end_in_contig), \
max(alignment.start_in_contig, alignment.end_in_contig)
for gene in contigs[alignment.name].genes:
if start_in_contig < gene.start < end_in_contig or start_in_contig < gene.end < end_in_contig:
corr_start = max(alignment.start, alignment.start + (gene.start - start_in_contig))
corr_end = min(alignment.end, alignment.end + (gene.end - end_in_contig))
gene_info = '{start:' + str(gene.start) + ',end:' + str(gene.end) + ',corr_start:' + \
str(corr_start) + ',corr_end:' + str(corr_end) + '}'
genes.append(gene_info)
data_str.append('{name:"' + alignment.name + '",corr_start:' + str(alignment.start) + ',corr_end:' +
str(alignment.end) + ',start:' + str(alignment.unshifted_start) + ',end:' +
str(alignment.unshifted_end) + ',misassemblies:"' + alignment.misassemblies + '",mis_ends:"' + misassembled_ends + '"')
if alignment.similar:
data_str[-1] += ',similar:"True"'
if alignment.ambiguous:
data_str[-1] += ',ambiguous:"True"'
if alignment.is_best_set:
data_str[-1] += ',is_best:"True"'
if contig_more_unaligned:
data_str[-1] += ',more_unaligned:"True"'
aligned_assemblies.add(alignment.label)
if overlapped_contigs[alignment]:
data_str.append(',overlaps:[ ')
data_str.append(','.join(overlapped_contigs[alignment]))
data_str.append(']')
if qconfig.gene_finding:
data_str.append(',genes:[' + ','.join(genes) + ']')
if ambiguity_alignments_by_labels and qconfig.ambiguity_usage == 'all':
data_str.append(',ambiguous_alignments:[ ')
data_str = add_contig_structure_data(data_str, ambiguity_alignments_by_labels[alignment.label][alignment.name],
ref_contigs, chr_full_names, contig_names_by_refs,
used_chromosomes, links_to_chromosomes, chr_names_by_id)
data_str[-1] = data_str[-1][:-1] + '],'
data_str[-1] = data_str[-1] + '},'
data_str[-1] = data_str[-1][:-1] + '];'
assembly_len = assemblies_len[assembly]
assembly_contigs = len(assemblies_contigs[assembly])
local_misassemblies = ms_types[assembly]['local'] // 2
ext_misassemblies = (sum(ms_types[assembly].values()) - ms_types[assembly]['interspecies translocation']) // 2 - \
local_misassemblies + ms_types[assembly]['interspecies translocation']
additional_assemblies_data += 'assemblies_len["' + assembly + '"] = ' + str(assembly_len) + ';\n'
additional_assemblies_data += 'assemblies_contigs["' + assembly + '"] = ' + str(assembly_contigs) + ';\n'
additional_assemblies_data += 'assemblies_misassemblies["' + assembly + '"] = "' + str(ext_misassemblies) + '+' + \
str(local_misassemblies) + '";\n'
if cov_data_str:
# adding coverage data
data_str.extend(cov_data_str)
if physical_cov_data_str:
data_str.extend(physical_cov_data_str)
if gc_data_str:
data_str.append('var gc_window_size = ' + (str(qconfig.GC_window_size_large) if
qconfig.large_genome else str(qconfig.GC_window_size)) + ';')
data_str.extend(gc_data_str)
misassemblies_types = ['relocation', 'translocation', 'inversion', 'interspecies translocation', 'local']
if not qconfig.is_combined_ref:
misassemblies_types.remove('interspecies translocation')
ms_counts_by_type = OrderedDict()
for ms_type in misassemblies_types:
factor = 1 if ms_type == 'interspecies translocation' else 2
ms_counts_by_type[ms_type] = sum(ms_types[assembly][ms_type] // factor for assembly in chr_to_aligned_blocks.keys())
total_ms_count = sum(ms_counts_by_type.values()) - ms_counts_by_type['local']
ms_selectors = []
for ms_type, ms_count in ms_counts_by_type.items():
ms_name = ms_type
if ms_count != 1 and ms_type != 'local':
ms_name += 's'
ms_selectors.append((ms_type, ms_name, str(ms_count)))
contigs_structure_str = get_contigs_structure(assemblies_contigs, chr_to_aligned_blocks, contigs_by_assemblies, ref_contigs, chr_full_names,
contig_names_by_refs, structures_by_labels, used_chromosomes, links_to_chromosomes, chr_names_by_id)
if contig_names_by_refs:
data_str.append(''.join(links_to_chromosomes))
data_str = '\n'.join(data_str)
return alignment_viewer_fpath, data_str, contigs_structure_str, additional_assemblies_data, ms_selectors, num_misassemblies, aligned_assemblies
def get_contigs_data(contigs_by_assemblies, nx_marks, assemblies_n50, structures_by_labels, contig_names_by_refs, ref_names,
chr_full_names):
additional_data = []
additional_data.append('var links_to_chromosomes;')
one_html = len(chr_full_names) == 1
if not one_html:
additional_data.append('links_to_chromosomes = {};')
for ref_name in ref_names:
chr_name = ref_name
if contig_names_by_refs and ref_name in contig_names_by_refs:
chr_name = contig_names_by_refs[ref_name]
chr_name = get_html_name(chr_name, chr_full_names)
additional_data.append('links_to_chromosomes["' + ref_name + '"] = "' + chr_name + '";')
contigs_sizes_str = ['var contig_data = {};']
contigs_sizes_str.append('var chromosome;')
contigs_sizes_lines = []
total_len = 0
min_contig_size = qconfig.min_contig
too_many_contigs = False
has_aligned_contigs = structures_by_labels.values()
for assembly in contigs_by_assemblies:
contigs_sizes_str.append('contig_data["' + assembly + '"] = [ ')
cum_length = 0
contigs = sorted(contigs_by_assemblies[assembly], key=lambda x: x.size, reverse=True)
last_contig_num = min(len(contigs), qconfig.max_contigs_num_for_size_viewer)
contig_threshold = contigs[last_contig_num - 1].size
not_used_nx = [nx for nx in nx_marks]
if len(contigs) > qconfig.max_contigs_num_for_size_viewer:
too_many_contigs = True
for i, alignment in enumerate(contigs):
if i >= last_contig_num:
break
cum_length, contigs_sizes_lines, align, not_used_nx = add_contig(cum_length, alignment, not_used_nx, assemblies_n50,
assembly, contigs, contigs_sizes_lines, i, structures_by_labels,
has_aligned_contigs=has_aligned_contigs)
contigs_sizes_str.append(align)
if len(contigs) > qconfig.max_contigs_num_for_size_viewer:
assembly_len = cum_length
remained_len = sum(alignment.size for alignment in contigs[last_contig_num:])
cum_length += remained_len
remained_genes = ['{start:' + str(gene.start) + ',end:' + str(gene.end) + '}' for contig in contigs[last_contig_num:]
for gene in contig.genes]
remained_contigs_name = str(len(contigs) - last_contig_num) + ' hidden contigs shorter than ' + str(contig_threshold) + \
' bp (total length: ' + format_long_numbers(remained_len) + ' bp)'
contigs_sizes_str.append(('{name:"' + remained_contigs_name + '", size:' + str(remained_len) +
', contig_type:"short_contigs"' + (',genes:[' + ','.join(remained_genes) + ']},'
if qconfig.gene_finding else '},')))
if not_used_nx and last_contig_num < len(contigs):
for i, alignment in enumerate(contigs[last_contig_num:]):
if not not_used_nx:
break
assembly_len, contigs_sizes_lines, align, not_used_nx = add_contig(assembly_len, alignment, not_used_nx, assemblies_n50,
assembly, contigs, contigs_sizes_lines, last_contig_num + i, structures_by_labels, only_nx=True)
total_len = max(total_len, cum_length)
contigs_sizes_str[-1] = contigs_sizes_str[-1][:-1] + '];\n\n'
contigs_sizes_str = '\n'.join(contigs_sizes_str)
contigs_sizes_str += 'var contigLines = [' + ','.join(contigs_sizes_lines) + '];\n\n'
contigs_sizes_str += 'var contigs_total_len = ' + str(total_len) + ';\n'
contigs_sizes_str += 'var minContigSize = ' + str(min_contig_size) + ';'
contig_viewer_data = contigs_sizes_str + '\n'.join(additional_data)
return contig_viewer_data, too_many_contigs
