def toClaRecord(self):
"""Return a Cla.Record"""
rec = Cla.Record()
rec.sid = self.sid
rec.residues = self.residues
rec.sccs = self.sccs
rec.sunid = self.sunid
n = self
while n.sunid != 0: #Not root node
rec.hierarchy.append((n.type, str(n.sunid)))
n = n.getParent()
rec.hierarchy.reverse()
return rec
def toClaRecord(self):
"""Return a Cla.Record"""
rec = Cla.Record()
rec.sid = self.sid
rec.residues = self.residues
rec.sccs = self.sccs
rec.sunid = self.sunid
n = self
while n.sunid != 0: #Not root node
rec.hierarchy[n.type] = str(n.sunid)
n = n.getParent()
# Order does not matter in the hierarchy field. For more info, see
# http://scop.mrc-lmb.cam.ac.uk/scop/release-notes.html
#rec.hierarchy.reverse()
return rec
def __init__(self,
cla_handle=None,
des_handle=None,
hie_handle=None,
dir_path=None,
db_handle=None,
version=None):
"""Build the SCOP hierarchy from the SCOP parsable files, or a sql backend.
If no file handles are given, then a Scop object with a single
empty root node is returned.
If a directory and version are given (with dir_path=.., version=...) or
file handles for each file, the whole scop tree will be built in memory.
If a MySQLdb database handle is given, the tree will be built as needed,
minimising construction times. To build the SQL database to the methods
write_xxx_sql to create the tables.
"""
self._sidDict = {}
self._sunidDict = {}
if cla_handle == des_handle == hie_handle == dir_path == db_handle == None:
return
if dir_path is None and db_handle is None:
if cla_handle == None or des_handle == None or hie_handle == None:
raise RuntimeError("Need CLA, DES and HIE files to build SCOP")
sunidDict = {}
self.db_handle = db_handle
try:
if db_handle:
# do nothing if we have a db handle, we'll do it all on the fly
pass
else:
# open SCOP parseable files
if dir_path:
if not version:
raise RuntimeError(
"Need SCOP version to find parsable files in directory"
)
if cla_handle or des_handle or hie_handle:
raise RuntimeError(
"Cannot specify SCOP directory and specific files")
cla_handle = _open_scop_file(dir_path, version, 'cla')
des_handle = _open_scop_file(dir_path, version, 'des')
hie_handle = _open_scop_file(dir_path, version, 'hie')
root = Node()
domains = []
root.sunid = 0
root.type = 'ro'
sunidDict[root.sunid] = root
self.root = root
root.description = 'SCOP Root'
# Build the rest of the nodes using the DES file
records = Des.parse(des_handle)
for record in records:
if record.nodetype == 'px':
n = Domain()
n.sid = record.name
domains.append(n)
else:
n = Node()
n.sunid = record.sunid
n.type = record.nodetype
n.sccs = record.sccs
n.description = record.description
sunidDict[n.sunid] = n
# Glue all of the Nodes together using the HIE file
records = Hie.parse(hie_handle)
for record in records:
if record.sunid not in sunidDict:
print record.sunid
n = sunidDict[record.sunid]
if record.parent != '': # Not root node
if record.parent not in sunidDict:
raise ValueError("Incomplete data?")
n.parent = sunidDict[record.parent]
for c in record.children:
if c not in sunidDict:
raise ValueError("Incomplete data?")
n.children.append(sunidDict[c])
# Fill in the gaps with information from the CLA file
sidDict = {}
records = Cla.parse(cla_handle)
for record in records:
n = sunidDict[record.sunid]
assert n.sccs == record.sccs
assert n.sid == record.sid
n.residues = record.residues
sidDict[n.sid] = n
# Clean up
self._sunidDict = sunidDict
self._sidDict = sidDict
self._domains = tuple(domains)
finally:
if dir_path:
# If we opened the files, we close the files
if cla_handle: cla_handle.close()
if des_handle: des_handle.close()
if hie_handle: hie_handle.close()
def __init__(self, cla_handle=None, des_handle=None, hie_handle=None,
dir_path=None, db_handle=None, version=None):
"""Build the SCOP hierarchy from the SCOP parsable files, or a sql backend.
If no file handles are given, then a Scop object with a single
empty root node is returned.
If a directory and version are given (with dir_path=.., version=...) or
file handles for each file, the whole scop tree will be built in memory.
If a MySQLdb database handle is given, the tree will be built as needed,
minimising construction times. To build the SQL database to the methods
write_xxx_sql to create the tables.
"""
self._sidDict = {}
self._sunidDict = {}
if cla_handle==des_handle==hie_handle==dir_path==db_handle==None: return
if dir_path is None and db_handle is None:
if cla_handle == None or des_handle==None or hie_handle==None:
raise RuntimeError("Need CLA, DES and HIE files to build SCOP")
sunidDict = {}
self.db_handle = db_handle
try:
if db_handle:
# do nothing if we have a db handle, we'll do it all on the fly
pass
else:
# open SCOP parseable files
if dir_path:
if not version:
raise RuntimeError("Need SCOP version to find parsable files in directory")
if cla_handle or des_handle or hie_handle:
raise RuntimeError("Cannot specify SCOP directory and specific files")
cla_handle = _open_scop_file( dir_path, version, 'cla')
des_handle = _open_scop_file( dir_path, version, 'des')
hie_handle = _open_scop_file( dir_path, version, 'hie')
root = Node()
domains = []
root.sunid=0
root.type='ro'
sunidDict[root.sunid] = root
self.root = root
root.description = 'SCOP Root'
# Build the rest of the nodes using the DES file
records = Des.parse(des_handle)
for record in records:
if record.nodetype =='px':
n = Domain()
n.sid = record.name
domains.append(n)
else :
n = Node()
n.sunid = record.sunid
n.type = record.nodetype
n.sccs = record.sccs
n.description = record.description
sunidDict[n.sunid] = n
# Glue all of the Nodes together using the HIE file
records = Hie.parse(hie_handle)
for record in records:
if record.sunid not in sunidDict:
print record.sunid
n = sunidDict[record.sunid]
if record.parent != '' : # Not root node
if record.parent not in sunidDict:
raise ValueError("Incomplete data?")
n.parent = sunidDict[record.parent]
for c in record.children:
if c not in sunidDict:
raise ValueError("Incomplete data?")
n.children.append(sunidDict[c])
# Fill in the gaps with information from the CLA file
sidDict = {}
records = Cla.parse(cla_handle)
for record in records:
n = sunidDict[record.sunid]
assert n.sccs == record.sccs
assert n.sid == record.sid
n.residues = record.residues
sidDict[n.sid] = n
# Clean up
self._sunidDict = sunidDict
self._sidDict = sidDict
self._domains = tuple(domains)
finally:
if dir_path:
# If we opened the files, we close the files
if cla_handle : cla_handle.close()
if des_handle : des_handle.close()
if hie_handle : hie_handle.close()