def load_sample_metadata(file_):
start = time.time()
fmd=snaputil.load_cpickle_file("%s.pkl" % (file_))
if fmd:
end = time.time()
taken = end-start
return fmd
start = time.time()
fmd={}
#dont need the hash-on-column headers just yet
with open(file_,"r") as f:
for line in f:
line = line.rstrip()
fields=line.split("\t")
fmd[fields[0]]=line
end = time.time()
taken = end-start
#sys.stderr.write("time taken to load samples from normal: %d\n" % taken)
snaputil.store_cpickle_file("%s.pkl" % (file_),fmd)
return fmd
def pack_sample_mapped_ids(line):
(sid, ids) = line.rstrip().split("|")
ba = bitarray()
i = 0
last_y = -1
ids_ = ids.split(",")
ids_.pop()
for y in sorted([int(x) for x in ids_]):
last_y = y
while i < y:
ba.append(False)
i += 1
ba.append(True)
i += 1
while i <= MAX_ID:
ba.append(False)
i += 1
#su.store_cpickle_file("./sample_ids1_%s.pkl" % (k),ba, compress=True)
su.store_cpickle_file("%s/%s.pkl" % (path, sid), ba, compress=False)
return True
def pack_sample_mapped_ids(line):
(sid,ids) = line.rstrip().split("|")
ba=bitarray()
i = 0
last_y = -1
ids_ = ids.split(",")
ids_.pop()
for y in sorted([int(x) for x in ids_]):
last_y = y
while i < y:
ba.append(False)
i+=1
ba.append(True)
i+=1
while i <= MAX_ID:
ba.append(False)
i+=1
#su.store_cpickle_file("./sample_ids1_%s.pkl" % (k),ba, compress=True)
su.store_cpickle_file("%s/%s.pkl" % (path,sid), ba, compress=False)
return True
def load_sample_metadata(file_):
start = time.time()
fmd=snaputil.load_cpickle_file("%s.pkl" % (file_))
if fmd:
end = time.time()
taken = end-start
return fmd
start = time.time()
fmd={}
#dont need the hash-on-column headers just yet
with open(file_,"r") as f:
for line in f:
line = line.rstrip()
fields=line.split("\t")
fmd[fields[0]]=line
end = time.time()
taken = end-start
#sys.stderr.write("time taken to load samples from normal: %d\n" % taken)
snaputil.store_cpickle_file("%s.pkl" % (file_),fmd)
return fmd
def __init__(self,
load_refseq=True,
load_canonical=True,
load_transcript=False):
self.ensembl_id_patt = re.compile('(ENST\d+)')
if load_refseq:
gene_file = "%s/%s" % (snapconf.TABIX_DB_PATH,
snapconf.REFSEQ_ANNOTATION)
gene_pickle_file = "%s.pkl" % (gene_file)
self.gene_map = snaputil.load_cpickle_file(gene_pickle_file)
if not self.gene_map:
self.load_gene_coords(gene_file)
snaputil.store_cpickle_file(gene_pickle_file, self.gene_map)
if load_canonical:
canonical_gene_file = "%s/%s" % (snapconf.TABIX_DB_PATH,
snapconf.CANONICAL_ANNOTATION)
canonical_gene_pickle_file = "%s.pkl" % (canonical_gene_file)
self.canonical_gene_map = snaputil.load_cpickle_file(
canonical_gene_pickle_file)
if not self.canonical_gene_map:
self.load_canonical_gene_coords(canonical_gene_file)
snaputil.store_cpickle_file(canonical_gene_pickle_file,
self.canonical_gene_map)
#per transcript exons
if load_transcript:
transcript_file = "%s/%s" % (snapconf.TABIX_DB_PATH,
snapconf.TABIX_GENE_INTERVAL_DB)
transcript_pickle_file = "%s.pkl" % (transcript_file)
self.transcript_map = snaputil.load_cpickle_file(
transcript_pickle_file)
if not self.transcript_map:
self.load_transcripts(transcript_file)
snaputil.store_cpickle_file(transcript_pickle_file,
self.transcript_map)
def __init__(self,load_refseq=True,load_canonical=True,load_transcript=False):
self.ensembl_id_patt = re.compile('(ENST\d+)')
if load_refseq:
gene_file = "%s/%s" % (snapconf.TABIX_DB_PATH,snapconf.REFSEQ_ANNOTATION)
gene_pickle_file = "%s.pkl" % (gene_file)
self.gene_map = snaputil.load_cpickle_file(gene_pickle_file)
if not self.gene_map:
self.load_gene_coords(gene_file)
snaputil.store_cpickle_file(gene_pickle_file,self.gene_map)
if load_canonical:
canonical_gene_file = "%s/%s" % (snapconf.TABIX_DB_PATH,snapconf.CANONICAL_ANNOTATION)
canonical_gene_pickle_file = "%s.pkl" % (canonical_gene_file)
self.canonical_gene_map = snaputil.load_cpickle_file(canonical_gene_pickle_file)
if not self.canonical_gene_map:
self.load_canonical_gene_coords(canonical_gene_file)
snaputil.store_cpickle_file(canonical_gene_pickle_file,self.canonical_gene_map)
#per transcript exons
if load_transcript:
transcript_file = "%s/%s" % (snapconf.TABIX_DB_PATH,snapconf.TABIX_GENE_INTERVAL_DB)
transcript_pickle_file = "%s.pkl" % (transcript_file)
self.transcript_map = snaputil.load_cpickle_file(transcript_pickle_file)
if not self.transcript_map:
self.load_transcripts(transcript_file)
snaputil.store_cpickle_file(transcript_pickle_file,self.transcript_map)
