def aai(self, options):
"""AAI command"""
check_file_exists(options.sorted_hit_table)
make_sure_path_exists(options.output_dir)
aai_calculator = AAICalculator(options.cpus)
aai_output_file, rbh_output_file = aai_calculator.run(
options.query_gene_file, None, options.sorted_hit_table,
options.evalue, options.per_identity, options.per_aln_len,
options.keep_rbhs, options.output_dir)
if rbh_output_file:
self.logger.info('Identified reciprocal best hits written to: %s' %
rbh_output_file)
self.logger.info('AAI between genomes written to: %s' %
aai_output_file)
def aai(self, options):
"""AAI command"""
check_file_exists(options.sorted_hit_table)
make_sure_path_exists(options.output_dir)
aai_calculator = AAICalculator(options.cpus)
aai_output_file, rbh_output_file = aai_calculator.run(options.query_gene_file,
None,
options.sorted_hit_table,
options.evalue,
options.per_identity,
options.per_aln_len,
options.keep_rbhs,
options.output_dir)
if rbh_output_file:
self.logger.info('Identified reciprocal best hits written to: %s' % rbh_output_file)
self.logger.info('AAI between genomes written to: %s' % aai_output_file)
def aai(self, options):
"""AAI command"""
self.logger.info('')
self.logger.info('*******************************************************************************')
self.logger.info(' [CompareM - aai] Calculating the AAI between homologs in genome pairs.')
self.logger.info('*******************************************************************************')
self.logger.info('')
check_dir_exists(options.rblast_dir)
make_sure_path_exists(options.output_dir)
genome_ids = []
protein_dir = os.path.join(options.rblast_dir, 'genes')
for f in os.listdir(protein_dir):
if f.endswith('.faa'):
genome_id = remove_extension(f, '.faa')
genome_ids.append(genome_id)
if not genome_ids:
self.logger.warning(' [Warning] No gene files found. Check the --protein_ext flag used to identify gene files.')
sys.exit()
aai_calculator = AAICalculator(options.cpus)
aai_calculator.run(genome_ids,
protein_dir,
options.rblast_dir,
options.per_identity,
options.per_aln_len,
options.write_shared_genes,
options.output_dir)
shared_genes_dir = os.path.join(options.output_dir, aai_calculator.shared_genes)
self.logger.info('')
self.logger.info(' Identified homologs between genome pairs written to: %s' % shared_genes_dir)
self.time_keeper.print_time_stamp()
def run(self, query_gene_file, target_gene_file, sorted_hit_table,
evalue_threshold, per_iden_threshold, per_aln_len_threshold,
num_top_targets, taxonomy_file, keep_rbhs, output_dir):
"""Classify genomes based on AAI to reference genomes.
Parameters
----------
query_gene_file : str
File with all query genes in FASTA format.
target_gene_file : str
File with all target genes in FASTA format.
sorted_hit_table : str
Sorted table indicating genes with sequence similarity.
evalue_threshold : float
Evalue threshold used to define a homologous gene.
per_identity_threshold : float
Percent identity threshold used to define a homologous gene.
per_aln_len_threshold : float
Alignment length threshold used to define a homologous gene.
num_top_targets : int
Number of top scoring target genomes to report per query genome.
taxonomy_file : str
File indicating taxonomic identification of all target genomes.
keep_rbhs : boolean
Flag indicating if RBH should be written to file.
output_dir : str
Directory to store AAI results.
"""
# read taxonomic identification of each genome
taxonomy = {}
if taxonomy_file:
for line in open(taxonomy_file):
genome_id, taxa_str = line.rstrip().split('\t')
taxonomy[genome_id] = taxa_str
# calculate AAI between query and target genomes
aai_output_dir = os.path.join(output_dir, 'aai')
make_sure_path_exists(aai_output_dir)
aai_calculator = AAICalculator(self.cpus)
aai_output_file, rbh_output_file = aai_calculator.run(
query_gene_file, target_gene_file, sorted_hit_table,
evalue_threshold, per_iden_threshold, per_aln_len_threshold,
keep_rbhs, aai_output_dir)
# determine matches to each query genomes
aai_results_file = os.path.join(aai_output_dir, 'aai_summary.tsv')
with open(aai_results_file) as f:
f.readline()
hits = defaultdict(list)
for line in f:
line_split = line.rstrip().split('\t')
query_id = line_split[0]
target_id = line_split[2]
aai = float(line_split[5])
of = float(line_split[7])
hits[query_id].append([target_id, aai, of])
# report top matches
results_file = os.path.join(output_dir, 'classify.tsv')
fout = open(results_file, 'w')
fout.write('Query Id\tTarget Id\tAAI\tOF\tScore')
if taxonomy:
fout.write('\tTarget Taxonomy')
fout.write('\n')
for query_id, cur_hits in hits.items():
cur_hits.sort(key=lambda x: x[1], reverse=True)
for i in xrange(0, min(num_top_targets, len(cur_hits))):
data = [query_id] + cur_hits[i]
fout.write('%s\t%s\t%.2f\t%.2f' % tuple(data))
aai = data[2]
of = data[3]
fout.write('\t%.2f' % (aai + of))
target_id = cur_hits[i][0]
if target_id in taxonomy:
fout.write('\t%s' % taxonomy[target_id])
fout.write('\n')
fout.close()
return results_file
def run(self, query_gene_file,
target_gene_file,
sorted_hit_table,
evalue_threshold,
per_iden_threshold,
per_aln_len_threshold,
num_top_targets,
taxonomy_file,
keep_rbhs,
output_dir):
"""Classify genomes based on AAI to reference genomes.
Parameters
----------
query_gene_file : str
File with all query genes in FASTA format.
target_gene_file : str
File with all target genes in FASTA format.
sorted_hit_table : str
Sorted table indicating genes with sequence similarity.
evalue_threshold : float
Evalue threshold used to define a homologous gene.
per_identity_threshold : float
Percent identity threshold used to define a homologous gene.
per_aln_len_threshold : float
Alignment length threshold used to define a homologous gene.
num_top_targets : int
Number of top scoring target genomes to report per query genome.
taxonomy_file : str
File indicating taxonomic identification of all target genomes.
keep_rbhs : boolean
Flag indicating if RBH should be written to file.
output_dir : str
Directory to store AAI results.
"""
# read taxonomic identification of each genome
taxonomy = {}
if taxonomy_file:
for line in open(taxonomy_file):
genome_id, taxa_str = line.rstrip().split('\t')
taxonomy[genome_id] = taxa_str
# calculate AAI between query and target genomes
aai_output_dir = os.path.join(output_dir, 'aai')
make_sure_path_exists(aai_output_dir)
aai_calculator = AAICalculator(self.cpus)
aai_output_file, rbh_output_file = aai_calculator.run(query_gene_file,
target_gene_file,
sorted_hit_table,
evalue_threshold,
per_iden_threshold,
per_aln_len_threshold,
keep_rbhs,
aai_output_dir)
# determine matches to each query genomes
aai_results_file = os.path.join(aai_output_dir, 'aai_summary.tsv')
with open(aai_results_file) as f:
f.readline()
hits = defaultdict(list)
for line in f:
line_split = line.rstrip().split('\t')
query_id = line_split[0]
target_id = line_split[2]
aai = float(line_split[5])
of = float(line_split[7])
hits[query_id].append([target_id, aai, of])
# report top matches
results_file = os.path.join(output_dir, 'classify.tsv')
fout = open(results_file, 'w')
fout.write('Query Id\tTarget Id\tAAI\tOF\tScore')
if taxonomy:
fout.write('\tTarget Taxonomy')
fout.write('\n')
for query_id, cur_hits in hits.items():
cur_hits.sort(key=lambda x: x[1], reverse=True)
for i in xrange(0, min(num_top_targets, len(cur_hits))):
data = [query_id] + cur_hits[i]
fout.write('%s\t%s\t%.2f\t%.2f' % tuple(data))
aai = data[2]
of = data[3]
fout.write('\t%.2f' % (aai+of))
target_id = cur_hits[i][0]
if target_id in taxonomy:
fout.write('\t%s' % taxonomy[target_id])
fout.write('\n')
fout.close()
return results_file