def predict_genes(index, contigs_fpath, gene_lengths, out_dirpath, tool_dirpath, tmp_dirpath, gmhmm_p_function,
prokaryote, num_threads):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
logger.info(' ' + qutils.index_to_str(index) + assembly_label)
err_fpath = os.path.join(out_dirpath, assembly_label + '_genemark.stderr')
genes = gmhmm_p_function(tool_dirpath, contigs_fpath, err_fpath, index, tmp_dirpath, num_threads)
if not genes:
unique_count = None
count = None # [None] * len(gene_lengths)
else:
tool_name = "genemark"
out_gff_fpath = os.path.join(out_dirpath, assembly_label + '_' + tool_name + '_genes.gff')
add_genes_to_gff(genes, out_gff_fpath, prokaryote)
if OUTPUT_FASTA:
out_fasta_fpath = os.path.join(out_dirpath, assembly_label + '_' + tool_name + '_genes.fasta')
add_genes_to_fasta(genes, out_fasta_fpath)
count = [sum([gene[3] - gene[2] > x for gene in genes]) for x in gene_lengths]
unique_count = len(set([gene[4] for gene in genes]))
total_count = len(genes)
logger.info(' ' + qutils.index_to_str(index) + ' Genes = ' + str(unique_count) + ' unique, ' + str(total_count) + ' total')
logger.info(' ' + qutils.index_to_str(index) + ' Predicted genes (GFF): ' + out_gff_fpath)
return unique_count, count
def run_gage(i, contigs_fpath, gage_results_dirpath, gage_tool_path, reference, tmp_dir):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
logger.info(' ' + qutils.index_to_str(i) + assembly_label + '...')
# run gage tool
log_out_fpath = os.path.join(gage_results_dirpath, 'gage_' + assembly_label + '.stdout')
log_err_fpath = os.path.join(gage_results_dirpath, 'gage_' + assembly_label + '.stderr')
logger.info(' ' + qutils.index_to_str(i) + 'Logging to files ' +
os.path.basename(log_out_fpath) + ' and ' +
os.path.basename(log_err_fpath) + '...')
log_out_f = open(log_out_fpath, 'w')
log_err_f = open(log_err_fpath, 'w')
return_code = qutils.call_subprocess(
['sh', gage_tool_path, reference, contigs_fpath, tmp_dir, str(qconfig.min_contig)],
stdout=log_out_f,
stderr=log_err_f,
indent=' ' + qutils.index_to_str(i),
only_if_debug=False)
if return_code != 0:
logger.info(' ' + qutils.index_to_str(i) + 'Failed.')
else:
logger.info(' ' + qutils.index_to_str(i) + 'Done.')
log_out_f.close()
log_err_f.close()
return return_code
def do(contigs_fpaths,
contig_report_fpath_pattern,
output_dirpath,
ref_fpath,
cov_fpath=None,
arcs=False,
similar=False,
coverage_hist=None):
make_output_dir(output_dirpath)
lists_of_aligned_blocks = []
total_genome_size = 0
reference_chromosomes = dict()
chr_names = []
for name, seq in fastaparser.read_fasta(ref_fpath):
chr_name = name.split()[0]
chr_names.append(chr_name)
chr_len = len(seq)
total_genome_size += chr_len
reference_chromosomes[chr_name] = chr_len
virtual_genome_shift = 100
sorted_ref_names = sorted(reference_chromosomes,
key=reference_chromosomes.get,
reverse=True)
sorted_ref_lengths = sorted(reference_chromosomes.values(), reverse=True)
cumulative_ref_lengths = [0]
for length in sorted(reference_chromosomes.values(), reverse=True):
cumulative_ref_lengths.append(cumulative_ref_lengths[-1] +
virtual_genome_shift + length)
virtual_genome_size = cumulative_ref_lengths[-1] - virtual_genome_shift
for contigs_fpath in contigs_fpaths:
report_fpath = contig_report_fpath_pattern % qutils.label_from_fpath_for_fname(
contigs_fpath)
aligned_blocks = parse_nucmer_contig_report(report_fpath,
sorted_ref_names,
cumulative_ref_lengths)
if aligned_blocks is None:
return None
for block in aligned_blocks:
block.label = qutils.name_from_fpath(contigs_fpath)
lists_of_aligned_blocks.append(aligned_blocks)
plot_fpath, assemblies = draw_alignment_plot(
contigs_fpaths, virtual_genome_size, sorted_ref_names,
sorted_ref_lengths, virtual_genome_shift, output_dirpath,
lists_of_aligned_blocks, arcs, similar, coverage_hist)
if assemblies and qconfig.create_contig_alignment_html:
js_data_gen(assemblies, contigs_fpaths, chr_names,
reference_chromosomes, output_dirpath, cov_fpath,
ref_fpath, virtual_genome_size)
return plot_fpath
def predict_genes(index, contigs_fpath, gene_lengths, out_dirpath, tool_dirpath, tmp_dirpath):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
logger.info(' ' + qutils.index_to_str(index) + assembly_label)
out_fpath = os.path.join(out_dirpath, assembly_label + '_glimmer')
err_fpath = os.path.join(out_dirpath, assembly_label + '_glimmer.stderr')
#out_gff_path, out_fasta_path, unique, total, cnt = glimmerHMM(tool_dir,
# fasta_path, out_path, gene_lengths, err_path)
out_gff_path, unique, total, cnt = glimmerHMM(tool_dirpath,
contigs_fpath, out_fpath, gene_lengths, err_fpath, tmp_dirpath, index)
if out_gff_path:
logger.info(' ' + qutils.index_to_str(index) + ' Genes = ' + str(unique) + ' unique, ' + str(total) + ' total')
logger.info(' ' + qutils.index_to_str(index) + ' Predicted genes (GFF): ' + out_gff_path)
return unique, cnt
def predict_genes(index, contigs_fpath, gene_lengths, out_dirpath,
tool_dirpath, tmp_dirpath, gmhmm_p_function, prokaryote,
num_threads):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
logger.info(' ' + qutils.index_to_str(index) + assembly_label)
err_fpath = os.path.join(out_dirpath, assembly_label + '_genemark.stderr')
genes = gmhmm_p_function(tool_dirpath, contigs_fpath, err_fpath, index,
tmp_dirpath, num_threads)
if not genes:
unique_count = None
count = None # [None] * len(gene_lengths)
else:
tool_name = "genemark"
out_gff_fpath = os.path.join(
out_dirpath, assembly_label + '_' + tool_name + '_genes.gff')
add_genes_to_gff(genes, out_gff_fpath, prokaryote)
if OUTPUT_FASTA:
out_fasta_fpath = os.path.join(
out_dirpath, assembly_label + '_' + tool_name + '_genes.fasta')
add_genes_to_fasta(genes, out_fasta_fpath)
count = [
sum([gene[3] - gene[2] > x for gene in genes])
for x in gene_lengths
]
unique_count = len(set([gene[4] for gene in genes]))
total_count = len(genes)
logger.info(' ' + qutils.index_to_str(index) + ' Genes = ' +
str(unique_count) + ' unique, ' + str(total_count) +
' total')
logger.info(' ' + qutils.index_to_str(index) +
' Predicted genes (GFF): ' + out_gff_fpath)
return unique_count, count
def do(contigs_fpaths, contig_report_fpath_pattern, output_dirpath,
ref_fpath, cov_fpath=None, arcs=False, similar=False, coverage_hist=None):
make_output_dir(output_dirpath)
lists_of_aligned_blocks = []
total_genome_size = 0
reference_chromosomes = dict()
chr_names = []
for name, seq in fastaparser.read_fasta(ref_fpath):
chr_name = name.split()[0]
chr_names.append(chr_name)
chr_len = len(seq)
total_genome_size += chr_len
reference_chromosomes[chr_name] = chr_len
virtual_genome_shift = 100
sorted_ref_names = sorted(reference_chromosomes, key=reference_chromosomes.get, reverse=True)
sorted_ref_lengths = sorted(reference_chromosomes.values(), reverse=True)
cumulative_ref_lengths = [0]
for length in sorted(reference_chromosomes.values(), reverse=True):
cumulative_ref_lengths.append(cumulative_ref_lengths[-1] + virtual_genome_shift + length)
virtual_genome_size = cumulative_ref_lengths[-1] - virtual_genome_shift
for contigs_fpath in contigs_fpaths:
report_fpath = contig_report_fpath_pattern % qutils.label_from_fpath_for_fname(contigs_fpath)
aligned_blocks = parse_nucmer_contig_report(report_fpath, sorted_ref_names, cumulative_ref_lengths)
if aligned_blocks is None:
return None
for block in aligned_blocks:
block.label = qutils.name_from_fpath(contigs_fpath)
lists_of_aligned_blocks.append(aligned_blocks)
plot_fpath, assemblies = draw_alignment_plot(
contigs_fpaths, virtual_genome_size, sorted_ref_names, sorted_ref_lengths, virtual_genome_shift, output_dirpath,
lists_of_aligned_blocks, arcs, similar, coverage_hist)
if assemblies and qconfig.create_contig_alignment_html:
js_data_gen(assemblies, contigs_fpaths, chr_names, reference_chromosomes, output_dirpath, cov_fpath, ref_fpath, virtual_genome_size)
return plot_fpath
def predict_genes(index, contigs_fpath, gene_lengths, out_dirpath,
tool_dirpath, tmp_dirpath):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
logger.info(' ' + qutils.index_to_str(index) + assembly_label)
out_fpath = os.path.join(out_dirpath, assembly_label + '_glimmer')
err_fpath = os.path.join(out_dirpath, assembly_label + '_glimmer.stderr')
#out_gff_path, out_fasta_path, unique, total, cnt = glimmerHMM(tool_dir,
# fasta_path, out_path, gene_lengths, err_path)
out_gff_path, unique, total, cnt = glimmerHMM(tool_dirpath, contigs_fpath,
out_fpath, gene_lengths,
err_fpath, tmp_dirpath,
index)
if out_gff_path:
logger.info(' ' + qutils.index_to_str(index) + ' Genes = ' +
str(unique) + ' unique, ' + str(total) + ' total')
logger.info(' ' + qutils.index_to_str(index) +
' Predicted genes (GFF): ' + out_gff_path)
return unique, cnt
def process_single_file(contigs_fpath, index, nucmer_path_dirpath,
genome_stats_dirpath, reference_chromosomes,
genes_container, operons_container):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
results = dict()
ref_lengths = {}
logger.info(' ' + qutils.index_to_str(index) + assembly_label)
nucmer_base_fpath = os.path.join(nucmer_path_dirpath,
assembly_label + '.coords')
if qconfig.use_all_alignments:
nucmer_fpath = nucmer_base_fpath
else:
nucmer_fpath = nucmer_base_fpath + '.filtered'
if not os.path.isfile(nucmer_fpath):
logger.error('Nucmer\'s coords file (' + nucmer_fpath +
') not found! Try to restart QUAST.',
indent=' ')
return None
coordfile = open(nucmer_fpath, 'r')
for line in coordfile:
if line.startswith('='):
break
# EXAMPLE:
# [S1] [E1] | [S2] [E2] | [LEN 1] [LEN 2] | [% IDY] | [TAGS]
#=====================================================================================
# 338980 339138 | 2298 2134 | 159 165 | 79.76 | gi|48994873|gb|U00096.2| NODE_0_length_6088
# 374145 374355 | 2306 2097 | 211 210 | 85.45 | gi|48994873|gb|U00096.2| NODE_0_length_6088
genome_mapping = {}
for chr_name, chr_len in reference_chromosomes.iteritems():
genome_mapping[chr_name] = [0] * (chr_len + 1)
contig_tuples = fastaparser.read_fasta(
contigs_fpath) # list of FASTA entries (in tuples: name, seq)
contig_tuples = sorted(contig_tuples,
key=lambda contig: len(contig[1]),
reverse=True)
sorted_contigs_names = [name for (name, seq) in contig_tuples]
genes_in_contigs = [0] * len(
sorted_contigs_names
) # for cumulative plots: i-th element is the number of genes in i-th contig
operons_in_contigs = [0] * len(sorted_contigs_names)
aligned_blocks_by_contig_name = {
} # for gene finding: contig_name --> list of AlignedBlock
for name in sorted_contigs_names:
aligned_blocks_by_contig_name[name] = []
for line in coordfile:
if line.strip() == '':
break
s1 = int(line.split('|')[0].split()[0])
e1 = int(line.split('|')[0].split()[1])
s2 = int(line.split('|')[1].split()[0])
e2 = int(line.split('|')[1].split()[1])
contig_name = line.split()[12].strip()
chr_name = line.split()[11].strip()
if chr_name not in genome_mapping:
logger.error("Something went wrong and chromosome names in your coords file (" + nucmer_base_fpath + ") " \
"differ from the names in the reference. Try to remove the file and restart QUAST.")
return None
aligned_blocks_by_contig_name[contig_name].append(
AlignedBlock(seqname=chr_name, start=s1, end=e1))
if s2 == 0 and e2 == 0: # special case: circular genome, contig starts on the end of a chromosome and ends in the beginning
for i in range(s1, len(genome_mapping[chr_name])):
genome_mapping[chr_name][i] = 1
for i in range(1, e1 + 1):
genome_mapping[chr_name][i] = 1
else: #if s1 <= e1:
for i in range(s1, e1 + 1):
genome_mapping[chr_name][i] = 1
coordfile.close()
# counting genome coverage and gaps number
covered_bp = 0
gaps_count = 0
gaps_fpath = os.path.join(genome_stats_dirpath,
assembly_label + '_gaps.txt')
gaps_file = open(gaps_fpath, 'w')
for chr_name, chr_len in reference_chromosomes.iteritems():
print >> gaps_file, chr_name
cur_gap_size = 0
aligned_len = 0
for i in range(1, chr_len + 1):
if genome_mapping[chr_name][i] == 1:
if cur_gap_size >= qconfig.min_gap_size:
gaps_count += 1
print >> gaps_file, i - cur_gap_size, i - 1
aligned_len += 1
covered_bp += 1
cur_gap_size = 0
else:
cur_gap_size += 1
ref_lengths[chr_name] = aligned_len
if cur_gap_size >= qconfig.min_gap_size:
gaps_count += 1
print >> gaps_file, chr_len - cur_gap_size + 1, chr_len
gaps_file.close()
results["covered_bp"] = covered_bp
results["gaps_count"] = gaps_count
# finding genes and operons
for container, feature_in_contigs, field, suffix in [
(genes_container, genes_in_contigs, reporting.Fields.GENES,
'_genes.txt'),
(operons_container, operons_in_contigs, reporting.Fields.OPERONS,
'_operons.txt')
]:
if not container.region_list:
results[field + "_full"] = None
results[field + "_partial"] = None
continue
total_full = 0
total_partial = 0
found_fpath = os.path.join(genome_stats_dirpath,
assembly_label + suffix)
found_file = open(found_fpath, 'w')
print >> found_file, '%s\t\t%s\t%s\t%s' % ('ID or #', 'Start', 'End',
'Type')
print >> found_file, '========================================='
# 0 - gene is not found,
# 1 - gene is found,
# 2 - part of gene is found
found_list = [0] * len(container.region_list)
for i, region in enumerate(container.region_list):
found_list[i] = 0
for contig_id, name in enumerate(sorted_contigs_names):
cur_feature_is_found = False
for cur_block in aligned_blocks_by_contig_name[name]:
if container.chr_names_dict[
region.seqname] != cur_block.seqname:
continue
# computing circular genomes
if cur_block.start > cur_block.end:
blocks = [
AlignedBlock(seqname=cur_block.seqname,
start=cur_block.start,
end=region.end + 1),
AlignedBlock(seqname=cur_block.seqname,
start=1,
end=cur_block.end)
]
else:
blocks = [cur_block]
for block in blocks:
if region.end <= block.start or block.end <= region.start:
continue
elif block.start <= region.start and region.end <= block.end:
if found_list[
i] == 2: # already found as partial gene
total_partial -= 1
found_list[i] = 1
total_full += 1
region_id = str(region.id)
if region_id == 'None':
region_id = '# ' + str(region.number + 1)
print >> found_file, '%s\t\t%d\t%d\tcomplete' % (
region_id, region.start, region.end)
feature_in_contigs[
contig_id] += 1 # inc number of found genes/operons in id-th contig
cur_feature_is_found = True
break
elif found_list[i] == 0 and min(
region.end, block.end) - max(
region.start,
block.start) >= qconfig.min_gene_overlap:
found_list[i] = 2
total_partial += 1
if cur_feature_is_found:
break
if cur_feature_is_found:
break
# adding info about partially found genes/operons
if found_list[i] == 2: # partial gene/operon
region_id = str(region.id)
if region_id == 'None':
region_id = '# ' + str(region.number + 1)
print >> found_file, '%s\t\t%d\t%d\tpartial' % (
region_id, region.start, region.end)
results[field + "_full"] = total_full
results[field + "_partial"] = total_partial
found_file.close()
logger.info(' ' + qutils.index_to_str(index) + 'Analysis is finished.')
return ref_lengths, (results, genes_in_contigs, operons_in_contigs)
def process_single_file(contigs_fpath, index, nucmer_path_dirpath, genome_stats_dirpath,
reference_chromosomes, genes_container, operons_container):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
results = dict()
ref_lengths = {}
logger.info(' ' + qutils.index_to_str(index) + assembly_label)
nucmer_base_fpath = os.path.join(nucmer_path_dirpath, assembly_label + '.coords')
if qconfig.use_all_alignments:
nucmer_fpath = nucmer_base_fpath
else:
nucmer_fpath = nucmer_base_fpath + '.filtered'
if not os.path.isfile(nucmer_fpath):
logger.error('Nucmer\'s coords file (' + nucmer_fpath + ') not found! Try to restart QUAST.',
indent=' ')
return None
coordfile = open(nucmer_fpath, 'r')
for line in coordfile:
if line.startswith('='):
break
# EXAMPLE:
# [S1] [E1] | [S2] [E2] | [LEN 1] [LEN 2] | [% IDY] | [TAGS]
#=====================================================================================
# 338980 339138 | 2298 2134 | 159 165 | 79.76 | gi|48994873|gb|U00096.2| NODE_0_length_6088
# 374145 374355 | 2306 2097 | 211 210 | 85.45 | gi|48994873|gb|U00096.2| NODE_0_length_6088
genome_mapping = {}
for chr_name, chr_len in reference_chromosomes.iteritems():
genome_mapping[chr_name] = [0] * (chr_len + 1)
contig_tuples = fastaparser.read_fasta(contigs_fpath) # list of FASTA entries (in tuples: name, seq)
contig_tuples = sorted(contig_tuples, key=lambda contig: len(contig[1]), reverse=True)
sorted_contigs_names = [name for (name, seq) in contig_tuples]
genes_in_contigs = [0] * len(sorted_contigs_names) # for cumulative plots: i-th element is the number of genes in i-th contig
operons_in_contigs = [0] * len(sorted_contigs_names)
aligned_blocks_by_contig_name = {} # for gene finding: contig_name --> list of AlignedBlock
for name in sorted_contigs_names:
aligned_blocks_by_contig_name[name] = []
for line in coordfile:
if line.strip() == '':
break
s1 = int(line.split('|')[0].split()[0])
e1 = int(line.split('|')[0].split()[1])
s2 = int(line.split('|')[1].split()[0])
e2 = int(line.split('|')[1].split()[1])
contig_name = line.split()[12].strip()
chr_name = line.split()[11].strip()
if chr_name not in genome_mapping:
logger.error("Something went wrong and chromosome names in your coords file (" + nucmer_base_fpath + ") " \
"differ from the names in the reference. Try to remove the file and restart QUAST.")
return None
aligned_blocks_by_contig_name[contig_name].append(AlignedBlock(seqname=chr_name, start=s1, end=e1))
if s2 == 0 and e2 == 0: # special case: circular genome, contig starts on the end of a chromosome and ends in the beginning
for i in range(s1, len(genome_mapping[chr_name])):
genome_mapping[chr_name][i] = 1
for i in range(1, e1 + 1):
genome_mapping[chr_name][i] = 1
else: #if s1 <= e1:
for i in range(s1, e1 + 1):
genome_mapping[chr_name][i] = 1
coordfile.close()
# counting genome coverage and gaps number
covered_bp = 0
gaps_count = 0
gaps_fpath = os.path.join(genome_stats_dirpath, assembly_label + '_gaps.txt')
gaps_file = open(gaps_fpath, 'w')
for chr_name, chr_len in reference_chromosomes.iteritems():
print >>gaps_file, chr_name
cur_gap_size = 0
aligned_len = 0
for i in range(1, chr_len + 1):
if genome_mapping[chr_name][i] == 1:
if cur_gap_size >= qconfig.min_gap_size:
gaps_count += 1
print >>gaps_file, i - cur_gap_size, i - 1
aligned_len += 1
covered_bp += 1
cur_gap_size = 0
else:
cur_gap_size += 1
ref_lengths[chr_name] = aligned_len
if cur_gap_size >= qconfig.min_gap_size:
gaps_count += 1
print >>gaps_file, chr_len - cur_gap_size + 1, chr_len
gaps_file.close()
results["covered_bp"] = covered_bp
results["gaps_count"] = gaps_count
# finding genes and operons
for container, feature_in_contigs, field, suffix in [
(genes_container,
genes_in_contigs,
reporting.Fields.GENES,
'_genes.txt'),
(operons_container,
operons_in_contigs,
reporting.Fields.OPERONS,
'_operons.txt')]:
if not container.region_list:
results[field + "_full"] = None
results[field + "_partial"] = None
continue
total_full = 0
total_partial = 0
found_fpath = os.path.join(genome_stats_dirpath, assembly_label + suffix)
found_file = open(found_fpath, 'w')
print >>found_file, '%s\t\t%s\t%s\t%s' % ('ID or #', 'Start', 'End', 'Type')
print >>found_file, '========================================='
# 0 - gene is not found,
# 1 - gene is found,
# 2 - part of gene is found
found_list = [0] * len(container.region_list)
for i, region in enumerate(container.region_list):
found_list[i] = 0
for contig_id, name in enumerate(sorted_contigs_names):
cur_feature_is_found = False
for cur_block in aligned_blocks_by_contig_name[name]:
if container.chr_names_dict[region.seqname] != cur_block.seqname:
continue
# computing circular genomes
if cur_block.start > cur_block.end:
blocks = [AlignedBlock(seqname=cur_block.seqname, start=cur_block.start, end=region.end + 1),
AlignedBlock(seqname=cur_block.seqname, start=1, end=cur_block.end)]
else:
blocks = [cur_block]
for block in blocks:
if region.end <= block.start or block.end <= region.start:
continue
elif block.start <= region.start and region.end <= block.end:
if found_list[i] == 2: # already found as partial gene
total_partial -= 1
found_list[i] = 1
total_full += 1
region_id = str(region.id)
if region_id == 'None':
region_id = '# ' + str(region.number + 1)
print >>found_file, '%s\t\t%d\t%d\tcomplete' % (region_id, region.start, region.end)
feature_in_contigs[contig_id] += 1 # inc number of found genes/operons in id-th contig
cur_feature_is_found = True
break
elif found_list[i] == 0 and min(region.end, block.end) - max(region.start, block.start) >= qconfig.min_gene_overlap:
found_list[i] = 2
total_partial += 1
if cur_feature_is_found:
break
if cur_feature_is_found:
break
# adding info about partially found genes/operons
if found_list[i] == 2: # partial gene/operon
region_id = str(region.id)
if region_id == 'None':
region_id = '# ' + str(region.number + 1)
print >>found_file, '%s\t\t%d\t%d\tpartial' % (region_id, region.start, region.end)
results[field + "_full"] = total_full
results[field + "_partial"] = total_partial
found_file.close()
logger.info(' ' + qutils.index_to_str(index) + 'Analysis is finished.')
return ref_lengths, (results, genes_in_contigs, operons_in_contigs)
def do(ref_fpath, contigs_fpaths, output_dirpath):
gage_results_dirpath = os.path.join(output_dirpath, 'gage')
# suffixes for files with report tables in plain text and tab separated formats
if not os.path.isdir(gage_results_dirpath):
os.mkdir(gage_results_dirpath)
########################################################################
gage_tool_path = os.path.join(qconfig.LIBS_LOCATION, 'gage', 'getCorrectnessStats.sh')
########################################################################
logger.print_timestamp()
logger.main_info('Running GAGE...')
metrics = ['Total units', 'Min', 'Max', 'N50', 'Genome Size', 'Assembly Size', 'Chaff bases',
'Missing Reference Bases', 'Missing Assembly Bases', 'Missing Assembly Contigs',
'Duplicated Reference Bases', 'Compressed Reference Bases', 'Bad Trim', 'Avg Idy', 'SNPs', 'Indels < 5bp',
'Indels >= 5', 'Inversions', 'Relocation', 'Translocation',
'Total units', 'BasesInFasta', 'Min', 'Max', 'N50']
metrics_in_reporting = [reporting.Fields.GAGE_NUMCONTIGS, reporting.Fields.GAGE_MINCONTIG, reporting.Fields.GAGE_MAXCONTIG,
reporting.Fields.GAGE_N50, reporting.Fields.GAGE_GENOMESIZE, reporting.Fields.GAGE_ASSEMBLY_SIZE,
reporting.Fields.GAGE_CHAFFBASES, reporting.Fields.GAGE_MISSINGREFBASES, reporting.Fields.GAGE_MISSINGASMBLYBASES,
reporting.Fields.GAGE_MISSINGASMBLYCONTIGS, reporting.Fields.GAGE_DUPREFBASES,
reporting.Fields.GAGE_COMPRESSEDREFBASES, reporting.Fields.GAGE_BADTRIM, reporting.Fields.GAGE_AVGIDY,
reporting.Fields.GAGE_SNPS, reporting.Fields.GAGE_SHORTINDELS, reporting.Fields.GAGE_LONGINDELS,
reporting.Fields.GAGE_INVERSIONS, reporting.Fields.GAGE_RELOCATION, reporting.Fields.GAGE_TRANSLOCATION,
reporting.Fields.GAGE_NUMCORCONTIGS, reporting.Fields.GAGE_CORASMBLYSIZE, reporting.Fields.GAGE_MINCORCONTIG,
reporting.Fields.GAGE_MAXCORCOTING, reporting.Fields.GAGE_CORN50]
tmp_dirpath = os.path.join(gage_results_dirpath, 'tmp')
if not os.path.exists(tmp_dirpath):
os.makedirs(tmp_dirpath)
n_jobs = min(len(contigs_fpaths), qconfig.max_threads)
from joblib import Parallel, delayed
return_codes = Parallel(n_jobs=n_jobs)(delayed(run_gage)(i, contigs_fpath, gage_results_dirpath, gage_tool_path, ref_fpath, tmp_dirpath)
for i, contigs_fpath in enumerate(contigs_fpaths))
if 0 not in return_codes:
logger.warning('Error occurred while GAGE was processing assemblies.'
' See GAGE error logs for details: %s' %
os.path.join(gage_results_dirpath, 'gage_*.stderr'))
return
## find metrics for total report:
for i, contigs_fpath in enumerate(contigs_fpaths):
assembly_label = qutils.label_from_fpath_for_fname(contigs_fpath)
report = reporting.get(contigs_fpath)
log_out_fpath = os.path.join(
gage_results_dirpath, 'gage_' + assembly_label + '.stdout')
logfile_out = open(log_out_fpath, 'r')
cur_metric_id = 0
for line in logfile_out:
if metrics[cur_metric_id] in line:
if (metrics[cur_metric_id].startswith('N50')):
report.add_field(metrics_in_reporting[cur_metric_id], line.split(metrics[cur_metric_id] + ':')[1].strip())
else:
report.add_field(metrics_in_reporting[cur_metric_id], line.split(':')[1].strip())
cur_metric_id += 1
if cur_metric_id == len(metrics):
break
logfile_out.close()
reporting.save_gage(output_dirpath)
if not qconfig.debug:
shutil.rmtree(tmp_dirpath)
logger.main_info('Done.')
