def output_pickle_file(self, results_output_dir, sample_label):
print("Serializing a total of %d events by Pickle." %(len(self.events)))
pickle_output_dir = os.path.join(results_output_dir, 'pickle')
if not os.path.isdir(pickle_output_dir):
os.mkdir(pickle_output_dir)
pickle_events_filename = os.path.join(pickle_output_dir, sample_label + '.pickle')
pickle_utils.write_pickled_file(self.events, pickle_events_filename)
return pickle_events_filename
def output_pickle_file(self, results_output_dir, sample_label): print "Serializing a total of %d events by Pickle." %(len(self.events)) pickle_output_dir = os.path.join(results_output_dir, 'pickle') if not os.path.isdir(pickle_output_dir): os.mkdir(pickle_output_dir) pickle_events_filename = os.path.join(pickle_output_dir, sample_label + '.pickle') pickle_utils.write_pickled_file(self.events, pickle_events_filename) return pickle_events_filename
def serialize_genes(gff_genes,
gff_filename,
output_dir,
compress_id=False):
"""
Output genes into pickle files by chromosome, by gene.
If asked, use compressed IDs (hashes) of the 'ID=' field in the GFF.
"""
genes_by_chrom = defaultdict(dict)
# Split up genes by chromosome
for gene_id, gene_info in gff_genes.iteritems():
gene_obj = gene_info["gene_object"]
gene_hierarchy = gene_info["hierarchy"]
genes_by_chrom[gene_obj.chrom][gene_id] = \
{'gene_object': gene_obj,
'hierarchy': gene_hierarchy}
if compress_id:
gene_compressed_id = compress_event_name(gene_id)
# Store compressed ID
genes_by_chrom[gene_obj.chrom][gene_id]['compressed_id'] \
= gene_compressed_id
# Mapping from gene IDs to pickled filename
gene_id_to_filename = {}
# Mapping from compressed IDs (hashes) to gene IDs
compressed_id_to_gene_id = {}
# Serialize all the genes in each chromosome into their
# own directory
for chrom, chrom_genes in genes_by_chrom.iteritems():
if chrom.startswith("chr"):
chrom_dir_name = chrom
else:
# Add chr-prefix for ease of finding directory
# in downstream steps.
chrom_dir_name = "chr%s" %(str(chrom))
# Make directory for chromosome if it doesn't already exist
chrom_dir = os.path.join(output_dir, chrom_dir_name)
if not os.path.isdir(chrom_dir):
#print "Making directory: %s" %(chrom_dir)
os.makedirs(chrom_dir)
t1 = time.time()
# Serialize each gene into a separate file
num_genes = len(genes_by_chrom[chrom])
for gene_id, gene_info in genes_by_chrom[chrom].iteritems():
gene_compressed_id = None
if compress_id:
gene_compressed_id = \
genes_by_chrom[chrom][gene_id]['compressed_id']
gene_filename = \
os.path.abspath(os.path.join(chrom_dir,
"%s.pickle" \
%(gene_compressed_id)))
else:
gene_filename = \
os.path.abspath(os.path.join(chrom_dir,
"%s.pickle" %(gene_id)))
# Write each gene/event's pickle file
pickle_utils.write_pickled_file({gene_id:
genes_by_chrom[chrom][gene_id]},
gene_filename)
# Record what filename was associated with this gene ID
gene_id_to_filename[gene_id] = gene_filename
# Record compressed ID (hash) to gene ID
if gene_compressed_id is not None:
compressed_id_to_gene_id[gene_compressed_id] = gene_id
t2 = time.time()
#print " - Chromosome serialization took %.2f seconds" %(t2 - t1)
print '.',
# Shelve the mapping from gene ids to filenames
shelved_filename = os.path.join(output_dir,
"genes_to_filenames.shelve")
shelved_data = shelve.open(shelved_filename)
for k, v in gene_id_to_filename.iteritems():
shelved_data[k] = v
shelved_data.close()
# Shelve the mapping from compressed gene ids to gene ids
shelved_filename = os.path.join(output_dir,
"compressed_ids_to_genes.shelve")
shelved_data = shelve.open(shelved_filename)
for k, v in compressed_id_to_gene_id.iteritems():
shelved_data[k] = v
shelved_data.close()
# Output a list of genes in ordinary GFF format
genes_filename = os.path.join(output_dir, "genes.gff")
#print "Outputting gene records in GFF format..."
#print " - Output file: %s" %(genes_filename)
with open(gff_filename) as gff_in:
with open(genes_filename, "w") as gff_out:
for line in gff_in:
if line.startswith("#"): continue
record_type = line.strip().split("\t")[2]
if record_type == "gene":
gff_out.write(line)
