def sortarchs(inputdir, outputdir):
archsdir = outputdir
Path.mkdir(archsdir)
sorted_archs = {}
loop_file_name = os.path.join(archsdir, 'ArchDB.{0}.db')
loop_split_file_name = os.path.join(archsdir, 'ArchDB.{0}.{1:02d}-{2:02d}.db')
sections_ini = [ 0, 4, 7,14,21]
sections_end = [ 4, 6,13,20, 0]
for archfile in Path.list_files(root = inputdir, pattern = '*.archObj'):
filename = os.path.basename(archfile)
data = filename.split('_')
length = int(data[0])
archtype = data[1]
sorted_archs.setdefault(archtype,{}).setdefault(length,[])
sorted_archs[archtype][length].append(archfile)
for archtype in sorted_archs:
SBIglobals.alert('verbose', None, "ARCHS: " + archtype + "\n")
fd = File(loop_file_name.format(archtype), 'w')
fdp = []
for x in range(len(sections_ini)):
fdp.append(File(loop_split_file_name.format(archtype, sections_ini[x], sections_end[x]), 'w'))
for length in sorted(sorted_archs[archtype]):
SBIglobals.alert('verbose', None, '\t{0}'.format(length))
for archfile in sorted_archs[archtype][length]:
SBIglobals.alert('verbose', None, '\t\t{0}\n'.format(archfile))
nsp = Arch.load(archfile)
fd.descriptor.write(nsp.archtype_format() + "\n")
for x in range(len(fdp)):
if length >= sections_ini[x] and (sections_end[x] == 0 or length <= sections_end[x]):
fdp[x].descriptor.write(nsp.archtype_format() + "\n")
fd.close()
for x in range(len(fdp)):
fdp[x].close()
def Arch2SQL(database, sqlfile, verbose):
if verbose:
sys.stderr.write("Retrieving data from {0} ...\n".format(database))
newsource = Source(name='PDBarch', source="http://www-pdb.org/")
outdir = os.path.join(os.path.join(os.path.abspath(sqlfile), '00'))
Path.mkdir(outdir)
sql_fd = gzip.open(os.path.join(outdir, '0000.sql.gz'), 'wb')
sql_fd.write(start_transaction())
sql_fd.write(newsource.toSQL())
sql_fd.write(end_transaction())
sql_fd.close()
files_list_by_pdb = {}
subdirs = ['archobj', 'superobj']
for subdir in subdirs:
for archobjfile in Path.list_files(os.path.join(database, subdir)):
if archobjfile.endswith('.archObj'):
data = tuple(
os.path.splitext(
os.path.split(archobjfile)[-1])[0].split('_')[2:])
files_list_by_pdb[data] = archobjfile
old_pdb = None
newArchSet = None
for dofdata in sorted(files_list_by_pdb):
pdb = dofdata[0] + '_' + dofdata[1]
if pdb != old_pdb:
if old_pdb is not None:
sql_fd.write(newArchSet.toSQL())
sql_fd.write(end_transaction())
sql_fd.close()
outdir = os.path.join(
os.path.join(os.path.abspath(sqlfile),
dofdata[0][1:3].lower()))
Path.mkdir(outdir)
if verbose:
sys.stderr.write("Retrieving loops from {0} ...\n".format(pdb))
sql_fd = gzip.open(os.path.join(outdir, pdb + '.sql.gz'), 'wb')
sql_fd.write(start_transaction())
if verbose:
sys.stderr.write("Printing data from {0} ...\n".format(pdb))
old_pdb = pdb
newArchSet = Arch(pdb)
newArchSet.archs = SSpair.load(files_list_by_pdb[dofdata])
sql_fd.write(newArchSet.toSQL())
sql_fd.write(end_transaction())
sql_fd.close()
if verbose:
sys.stderr.write("End execution.\n")
def get_PDB(self, pdbID):
if self.has_local:
rootdir = os.path.join(self.local, pdbID.lower()[1:3])
for pdb_file in Path.list_files(root=rootdir, pattern='*.ent.gz'):
newfile = File(file_name=pdb_file, action='r')
if newfile.prefix.lstrip('pdb').upper() == pdbID.upper():
return pdb_file
#If we do not find it in local (or we do not have a local) we search it on the FTP
pdb_file = 'pdb' + pdbID.lower() + '.ent.gz'
source = 'ftp://' + PDBftp['address'] + os.path.join(
PDBftp['structures'], pdbID[1:3].lower(), pdb_file)
try:
urllib.urlretrieve(source, pdb_file)
except:
return False
return os.path.abspath(pdb_file)
def _process(self):
tmoFile = File(self._pdbtmfile, 'w', True)
for xmlfile in Path.list_files(
os.path.join(self._local, 'pdbtm/database/'), '*.xml'):
xmldata = TM(
pdb=os.path.splitext(os.path.split(xmlfile)[1])[0].upper())
skip_chains = set()
read = False
fdxml = open(xmlfile)
for line in fdxml:
if line.startswith(' <TMRES>'): xmldata.tmres = line
elif line.startswith(' <TMTYPE'): xmldata.tmtype = line
elif line.startswith(' <PDBKWRES'): xmldata.kwres = line
elif line.startswith(' <SIDEDEFINITION'):
m = re.search('Side1="(\S+)"', line)
xmldata.side = m.group(1)
elif line.startswith(' <APPLY_TO_CHAIN'):
m = re.search('NEW_CHAINID=\"(\S{1})\"', line)
if m: skip_chains.add(m.group(1))
elif line.startswith(' <CHAIN '):
m = re.search(
'CHAINID=\"(\S{1})\" NUM_TM=\"(\d{1})\" TYPE=\"(\S+)\"',
line)
if m:
chain, num, tmtype = m.group(1), m.group(2), m.group(3)
if not chain in skip_chains:
cdata = tuple([chain, num, tmtype])
xmldata.set_chain(cdata)
read = True
elif line.startswith(' <REGION ') and read:
m = re.search(
'pdb_beg=\"(\-*\d+\w*)\"[\s\S]+pdb_end=\"(\-*\d+\w*)\"\s+type=\"(\w{1})\"',
line)
ini, end, tmtype = m.group(1), m.group(2), m.group(3)
xmldata.set_chain(cdata, tuple([ini, end, tmtype]))
elif line.startswith(' </CHAIN>'):
read = False
fdxml.close()
if len(xmldata.chains) > 0:
tmoFile.write(str(xmldata) + "\n")
tmoFile.close()
def DS2SQL(database, looplist, sqlfile, verbose):
Path.mkdir(sqlfile)
for dsfile in Path.list_files(database):
subclasstype = os.path.split(dsfile)[-1].split('.')[1]
classification = Cclass(subclasstype)
if verbose:
sys.stderr.write(
"Retrieving data for subclass {0} ...\n".format(subclasstype))
loops = readlist(looplist, subclasstype)
sql_fd = gzip.open(os.path.join(sqlfile, subclasstype + '.sql.gz'),
'wb')
sql_fd.write(start_transaction())
sql_in = open(dsfile)
read = False
for line in sql_in:
dataline = line.rstrip('\n')
#SKIP LINES
if line.startswith('==') or line.startswith('***') or len(
line.strip()) == 0 or line.startswith(
'---- P R O T E I N C O D E ----'):
continue
if line.startswith('CONSENSUS & MULTIPLE ALIGNEMENT IN THE'):
data = line.split(':')[-1].strip().split()
classification.subclasses = Subclass(
tuple([data[0].strip(), data[3].strip()]), data[4])
workscls = classification.lastsubclass
read = True
continue
if line.startswith('GLOBAL STATISTICS'):
read = False
continue
if read:
if line.startswith(
' SEQUENCE ALIGNEMENT :'
):
parse_mode, counter = 'P', 0
elif line.startswith(
' ACCESSIBLE SURFACE ALIGNEMENT :'
):
parse_mode, counter = 'E', 0
elif line.startswith(
' RAMACHANDRAN :'
):
parse_mode, counter = 'R', 0
elif line.startswith(
' SECONDARY STRUCTURE :'
):
parse_mode, counter = 'S', 0
elif line.startswith('--------- CONSENSUS THORNTON :'):
workscls.add_consensus(dataline, 'DS', loops)
elif line.startswith('--------- CONSENSUS TOPOLOGY'):
workscls.add_topology(dataline, 'DS')
elif line.startswith('CENTROIDE POLAR COORD. :'):
workscls.add_coordinates(dataline)
elif line.startswith('--------- RAMACHANDRAN PATTERN :'):
workscls.ram_pat = re.sub(
'\(X\)', '',
dataline.split(':')[1].strip().strip('.'))
elif line.startswith('--------- SEQUENCE PATTERN :'):
workscls.seq_pat = re.sub(
'\(X\)', '',
dataline.split(':')[1].strip().strip('.'))
elif line.startswith('--------- BURIAL PATTERN :'):
workscls.exp_pat = re.sub(
'\(X\)', '',
dataline.split(':')[1].strip().strip('.'))
elif line.startswith(' '
) and len(dataline) < 400:
if parse_mode == 'P': workscls.loops = Loop(info=dataline)
if parse_mode == 'E':
workscls.loops[counter].add_surface(info=dataline)
counter += 1
if parse_mode == 'R':
workscls.loops[counter].add_ramachandran(info=dataline)
counter += 1
if parse_mode == 'S':
workscls.loops[counter].add_secondary_str(
info=dataline)
counter += 1
sql_fd.write(classification.toSQL('DS'))
sql_in.close()
sql_fd.write(end_transaction())
sql_fd.close()
if verbose: sys.stderr.write("End execution.\n")
def localPDBeChems(self):
for chem_file in Path.list_files(root = self.local, pattern = '*.cif'):
yield chem_file
def localPDBs(self):
for pdb_file in Path.list_files(root=self.local, pattern='*.ent.gz'):
yield pdb_file