def execute_query(self, query_file, blast_output_file = None,
work_directory = os.getcwd()):
'''
Execute BLAST given a query sequence.
@param: query_file
@pdef: Fasta file with the query sequence.
@pdefault: 'QuerySequence'
@ptype: {String} or {File} or {Fasta}
@param: blast_output_file
@pdef: name of the temporary BLAST output file.
@pdefault: query_file.prefix + job.pid + .blast.xml.out
@ptype: {String}
@param: work_directory
@pdef: Directory to which the temporary files will be created.
@pdefault: Current working directory.
@ptype: {String}
@raises: {AttributeError} if query_file is multi-fasta.
@raises: {BlastError} in BLAST execution or output parsing errors.
@returns: {BlastResult}
'''
if isinstance(query_file, basestring) or isinstance(query_file, File):
newFasta = Fasta(fasta_file = query_file)
elif isinstance(query_file, Fasta):
newFasta = query_file
if newFasta.is_multifasta:
msg = 'Blasts can only be executed one at a time due to XML output restrictions.'
raise AttributeError(msg)
# All the sequence is unknown, it will crash blast
newFasta.load()
query_sequence = newFasta.sequence
if len(re.sub(r'[Xx]', '', query_sequence.sequence)) == 0:
SBIg.warn(self, 'Created an empty BlastResult.')
return BlastResult(query_name = query_sequence.id,
query_sequence = query_sequence.sequence)
Path.mkdir(work_directory)
file_prefixes = ".".join([newFasta.file.prefix, str(os.getpid())])
file_prefixes = os.path.join(work_directory, file_prefixes)
tmp_output = file_prefixes + ".blast.xml.out"
tmp_output = tmp_output if blast_output_file is None else os.path.join(work_directory, blast_output_file)
self._execute(input_file = newFasta, output_file = tmp_output)
blast_result = self._parse_blast(newFasta.sequence.sequence, tmp_output)
self._clean([tmp_output, ])
return blast_result
def _add_AAc(self, chain):
if self._AAdo and chain.chaintype == 'P':
raise NotImplementedError
ppi_id = PPInnerContact.test_identifier(chain)
SBIglobals.alert('debug', self, 'Analyzing Protein Inner Contacts {0} for {1.chain}'.format(ppi_id, chain))
ppi = PPInnerContact(chain, self._AA_type, self._AA_distance)
SBIglobals.alert('debug', self, '\tAdding new Inner Contacts')
self._AAcontacts[ppi_id] = ppi
def get_residue_by_identifier(self, identifier):
if len(self._chaindict) == 0:
for r in self.full_chain:
self._chaindict[r.identifier.strip()] = r
try:
r = self._chaindict[str(identifier.strip())]
except KeyError as e:
SBIglobals.alert('error', self, '{0} is not a valid identifier for chain {1}'.format(identifier, self.globalID), e)
else:
return r
def index_file(self, value):
'''
@param: value
@pdef: name of the related index file
@ptype: {String}
'''
if os.path.isfile(value):
self._index_file = value
else:
SBIg.warn(self, '{0} does not exist.'.format(value))
def _check_action(self, action):
if not action in self.available_action: # Raise 'Wrong action' Error
raise File(1, action, self.available_action)
if self.is_gziped and not action.endswith('b'):
action += 'b' # Better if working with compressed files
self._action = action
SBIglobals.alert('debug', self,
'\tAction {0} is OK...'.format(self._action))
def _build(self):
count = 1
SBIglobals.alert('debug', self, 'Analyzing Inner Contacts of {0:03} chains'.format(len(self.pdb)))
for chain in self.pdb.chains:
SBIglobals.alert('debug', self, 'Analyzing Chain {0:03} out of {1:03}'.format(count, len(self.pdb)))
self._add_AAc(chain)
self._add_NCc(chain)
self._add_HTc(chain)
count += 1
def _execute(self):
'''
Executes the DSSP call.
'''
self._EXE.add_parameter(self._pdbfile)
self._EXE.add_parameter(self._dsspfile)
try:
self._EXE.execute(silent=True)
except SystemError, e:
msg = 'Some error occurred while executing dssp\n{0}\n'.format(e)
SBIg.throw(self, msg, e)
def is_reference_to_db(self, db_minicode):
db_minicode = db_minicode.upper()
if not db_minicode in self.valid_references:
line1 = '{0} is not a valid DBref code'.format(db_minicode)
line2 = 'Available codes: {0}'.format(self.valid_references.keys())
SBIglobals.error(self, "\n".join([line1, line2]))
#SWS and TREMBL are parts of UNP
if db_minicode == 'UNP' and self._db in ['SWS', 'TREMBL']:
return True
return self._db == db_minicode
def dump(self, object_file, overwrite = None):
"""
- dump(): Stores the object into a file
- object_file (string): Name for the output file
@Mandatory
- overwrite (bool): Overwrite previous file of the same name
@Raises FileError
"""
SBIglobals.alert('verbose', self, 'Writting object to file {0}'.format(object_file))
dumpFile = File(file_name = object_file, action = 'wb', overwrite = overwrite)
pickle.dump(self, dumpFile.descriptor)
dumpFile.close()
def _add_HTc(self, chain):
if self._HTdo and chain.chaintype == 'P':
phi_id = PHInnerContact.test_identifier(chain)
SBIglobals.alert(
'debug', self,
'Analyzing Protein-Heteroatom Inner Contacts {0} for {1.chain}'
.format(phi_id, chain))
phi = PHInnerContact(chain, self._HT_type, self._HT_distance)
SBIglobals.alert('debug', self, '\tAdding new Inner Contacts')
self._HTcontacts[phi_id] = phi
def load(object_file):
"""
> load(): Retrieves the object from a python object file
- object_file (string): Name of the file containing the object
@Mandatory
@Returns the loaded object.
@staticmethod: can be called without any instance declared
@Raises FileError
"""
SBIglobals.alert('verbose', StorableObject, 'Preparing to load object from file {0}'.format(object_file))
Object = None
loadFile = File(file_name = object_file, action='rb')
Object = pickle.load(loadFile.descriptor)
loadFile.close()
return Object
def _check_file(self):
if self._action.startswith('r'):
if not os.path.isfile(self.full):
raise FileError(3, self.full, 'noexists')
if not os.access(self.full, os.R_OK):
raise FileError(4, self.full, 'read')
if self._action.startswith('w') or self._action.startswith('a'):
if os.path.isfile(self.full):
if not self._overwrite: raise FileError(3, self.full, 'exists')
if not os.path.isdir(self.dir):
raise FileError(4, self.dir, 'nodir')
if not os.access(self.dir, os.W_OK):
raise FileError(4, self.dir, 'write')
SBIglobals.alert('debug', self, '\tFile is OK...')
def __init__(self, file_name=None, action='r', overwrite=None):
if file_name is None:
raise FileError(0) # Raise 'No file specified' Error
self._file = file_name
SBIglobals.alert('debug', self,
'Preparing File: {0}'.format(self.full))
self._action = None
self._check_action(action.lower()) #action must be valid
self._fd = None
# Local overwrite takes precedence over Global overwrite
self._overwrite = SBIglobals.decide_overwrite(overwrite)
self._check_file()
def __init__(self, database, overwrite=None, clean=True):
'''
@param: database
@pdef: database to blast upon.
@ptype: {String}
@param: overwrite
@pdef: For writing actions. Decides whether it can overwrite an
existing file.
@pdefault: _SBIglobals.overwrite_
@ptype: {Boolean}
@param: clean
@pdef: remove the temporary files after the data is read.
@pdefault: _True_
@pclash: if _SBIglobals.debug_ is _True_, clean is _False_
@ptype: {Boolean}
@raises: {HmmError}
'''
self._error = HmmError()
#hmmer executable configuration
if HmmExe._EXE is None:
self._set_default_executable('hmmer')
HmmExe._DBFORMATER = HmmExe._CONFIG.get('hmmer', 'dbformatexe')
# Local overwrite takes precedence over Global overwrite
self._overwrite = SBIg.decide_overwrite(overwrite)
self._database = self._check_database(os.path.abspath(database))
self._clean_files = clean
#Optional execution parameters
self._parameters = {'attr': [], 'flag': []}
def items(self):
'''
Loops through the items of the database
@yields: Object depending on the database.
'''
if not self.has_local:
SBIg.throw(self, 'A local database needs to be build first',
IOError)
for ifile in self._ITEM_FILES:
ifile = os.path.join(self.local, ifile)
f = File(ifile)
for line in f.read():
yield self._DBOBJECT.grab(line.strip())
f.close()
def __init__(self, pdb_line):
'''
@type pdb_line: String
@param pdb_pdb_line: Line of a PDB. Starts with DBREF
@raise AttributeError if line does not start with DBREF
'''
if not pdb_line.startswith('DBREF '):
SBIglobals.error(self, '{0} cannot create DBref'.format(pdb_line))
data = self._process_line(pdb_line)
self._pdb = data[0]
self._chain = data[1]
self._start = data[2]
self._end = data[3]
self._db = data[4]
self._ref = data[5]
def create(self):
'''
Create the local database.
Returns _True_ on success, _False_ otherwise
@return: {Boolean}
'''
if self.has_local:
SBIg.warn(self, 'A local copy exist. Executing update.')
return self.update()
self._download_sources()
self._process()
self._save_release()
# self._clean_sources()
return True
def retrieve(self, sequence_ids, all_but=False, prefix_size=None):
'''
Get specific sequences from the FASTA file.
@param: sequence_ids
@pdef: sequence identifier(s)
@ptype: {String}, {List} or {Set}
@param: all_but
@pdef: Flag. Instead of retrieving the given ids, we retrieve all
except the given ids.
@pdefault: _False_
@ptype: {Boolean}
@param: prefix_size
@pdef: maximum characters for the prefix. If _None_, all the
characters are included.
@pdefault: _None_
@ptype: {Integer}
@raises: {AttributeError} if sequence_ids is not a valid type.
@return: {List} of {Sequence}
'''
info = 'Skipping sequence {0}' if all_but else 'Getting sequence {0}'
if not isinstance(sequence_ids, (list, set)):
SBIg.alert('debug', self, info.format(sequence_ids))
else:
SBIg.alert('debug', self, [info.format(x) for x in sequence_ids])
if isinstance(sequence_ids, basestring):
sequence_ids = set([sequence_ids])
if isinstance(sequence_ids, list):
sequence_ids = set(sequence_ids)
if isinstance(sequence_ids, set):
sequences = []
for s in self.live_show():
seq_id = s.id if prefix_size is None else s.id[:prefix_size]
if seq_id in sequence_ids and not all_but:
sequences.append(s)
if seq_id not in sequence_ids and all_but:
sequences.append(s)
return sequences
else:
raise AttributeError('sequence_ids must be a string, list or set.')
def _format_database(self, database):
'''
Executes the blast script to format the database.
@param: database
@pdef: database to blast upon.
@ptype: {String}
'''
SBIg.warn(self, 'Formating {0} for blast.'.format(database))
dbexe = Executable(executable = BlastExe._DBFORMATER,
path = self._EXE.path)
dbexe.add_attribute(database, '-in')
dbexe.add_attribute(self._search_type[0:4], '-dbtype')
SBIg.alert('debug', self, 'Executing command {0}\n'.format(dbexe))
dbexe.execute()
def _build(self):
if len(self.protein.aminoacids) == 0 or len(
self.nucleotide.nucleotides) == 0:
return
super(PNInterface, self)._build()
for i in range(len(self._filtered[0])):
SBIglobals.alert(
'deepdebug', self,
'Analyze AN Contact for {0.type}:{0.number} - {1.type}:{1.number}'
.format(self.protein.aminoacids[self._filtered[0][i]],
self.nucleotide.nucleotides[self._filtered[1][i]]))
new_contact = ContactAN(
aminoacid=self.protein.aminoacids[self._filtered[0][i]],
nucleotide=self.nucleotide.nucleotides[self._filtered[1][i]],
threshold_type=self.threshold_type,
threshold_distance=self.threshold_distance)
if new_contact.is_underthreshold:
self.contacts = new_contact
def list_directories(root=os.curdir, rootless=False):
"""
> list_directories(): Returns all dirs in a directory tree
- root (string): Root of the directory tree to search
@Default: current working directory
- rootless (bool) : When False, the name of the dirs are returned with
absolute path. Otherwise the root is removed.
@Default: False
@Yields directory names
"""
for path, dirs, files in os.walk(os.path.abspath(root)):
for onedir in dirs:
SBIglobals.alert(
'debug', Path(),
'Found directory {0}'.format(os.path.join(path, onedir)))
if not rootless:
yield os.path.join(path, onedir)
else:
yield os.path.join(path, onedir).replace(root, '')
def _process(self, update=False):
'''
Transform the source files into the final local db files.
@param: update
@pdef: toggles between create and update processing
@pdefault: _False_
@ptype: {Boolean
'''
if update:
old = self._RELEASE['total_items'].copy()
j = 0
for i in range(len(self._SOURCES)):
dfilen = os.path.join(self.local, self._SOURCES[i])
ofilen = os.path.join(self.local, self._MANDATORY_FILES[j])
ffilen = os.path.join(self.local, self._MANDATORY_FILES[j + 1])
if not os.path.isfile(dfilen):
continue
SBIg.alert('verbose', self, 'Parsing: {0}'.format(dfilen))
SBIg.alert('verbose', self, 'DB file to: {0}'.format(ofilen))
SBIg.alert('verbose', self, 'Fasta file to: {0}'.format(ffilen))
dfile = File(dfilen)
ofile = File(ofilen, 'w', update)
ffile = File(ffilen, 'w', update)
protein = None
for protein in Connect._parse_uniprot(dfile):
pname = protein.entry_name
pvers = protein.version
SBIg.alert('verbose', self, 'Protein: {0}'.format(pname))
if not update:
self._RELEASE['total_items'][pname] = pvers
else:
if pname not in self._RELEASE['total_items']:
self._RELEASE['new_items'][pname] = pvers
else:
del (old[pname])
if self._RELEASE['total_items'][pname] != pvers:
self._RELEASE['update_items'][pname] = pvers
ffile.write(protein.sequence.format('FASTA') + '\n')
ofile.write(protein.json() + '\n')
j += 2
dfile.close()
ofile.close()
ffile.close()
if update:
self._RELEASE['total_items'].update(self._RELEASE['new_items'])
self._RELEASE['total_items'].update(self._RELEASE['update_items'])
self._RELEASE['deleted_items'] = old
for k in self._RELEASE['deleted_items']:
del (self._RELEASE['total_items'][k])
def _format_database(self, database):
'''
Executes the hmmer script to format the database.
@param: database
@pdef: database to blast upon.
@ptype: {String}
'''
SBIg.warn(self, 'Formating {0} for hmmer.'.format(database))
dbexe = Executable(executable=HmmExe._DBFORMATER, path=self._EXE.path)
if self.overwrite:
dbexe.add_parameter('-f')
dbexe.add_parameter(database)
SBIg.alert('debug', self, 'Executing command {0}\n'.format(dbexe))
dbexe.execute()
def reposition(self, matrix=None, vector=None):
"""
Rotates and translates the {Residue} according to a matrix and translational vector
@type matrix: numpy.matrix
@type vector: numpy.array
"""
if matrix is None: matrix = np.identity(3, float)
if vector is None: vector = np.zeros(3, float)
SBIglobals.alert('deepdebug', self,
'Reposition residue {0.type}:{0.number}'.format(self))
self._backbone_coordinates = None
self._sidechain_coordinates = None
for atom in self.atoms:
SBIglobals.alert('deepdebug', self,
'Atom {0.name} {0.is_backbone}'.format(atom))
atom.rotate(matrix=matrix)
atom.translate(vector=vector)
self._add_to_matrix(atom)
def __init__(self, pdb_file=None):
SBIglobals.alert('debug', self,
'Loading PDB file {0}'.format(pdb_file))
super(PDB, self).__init__(pdb_file=pdb_file,
dehydrate=True,
header=True)
SBIglobals.alert(
'debug', self,
'Calculating Inner Contacts for Protein - Heteroatom')
self.innercontacts = InnerContacts(pdb=self,
AA=False,
NC=False,
HT=True,
HT_type="min",
HT_distance=6)
SBIglobals.alert(
'debug', self,
'Calculating PP and PN interfaces of the biomolecules in the PDB')
self.interfaces = Complex(pdb=self,
biomolecule=True,
PPI=True,
PPI_type="cb",
PPI_distance=12,
PNI=True,
PNI_type="min",
PNI_distance=8,
PHI=True,
PHI_type="min",
PHI_distance=6)
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 _download_sources(self):
'''
Download the source files to local directory.
'''
for dfile in self._SOURCES:
download = False
source = os.path.join(self._FTP, dfile)
source_size = DBlink._file_size(source)
# source_date = DBlink._file_date(source)
DBlink._SOURCES_SIZES.append(source_size)
destination = os.path.join(self.local, dfile)
if not os.path.isfile(destination):
if DBlink._RELEASE['date'] != DBlink._TODAY:
# if source_date < DBlink._RELEASE:
# SBIg.alert('verbose', self, 'No new updates in the source side' +
# ' for {0}'.format(source) +
# ' since the last local update.')
# else:
download = True
else:
download = True
else:
SBIg.alert(
'verbose', self,
'Looks like {0} has already been downloaded.'.format(
source))
if download:
SBIg.alert(
'verbose', self,
'Downloading {0} to {1}'.format(source, destination))
SBIg.alert(
'verbose', self,
'Source file size is {0:.3f} MB.'.format(source_size))
urllib.urlretrieve(source, destination)
def list_files(root=os.curdir,
pattern='*',
avoid_empty_files=True,
rootless=False):
"""
> list_files() : Returns any file in a directory tree matching a specific pattern
- root (string): Root of the directory tree to search
@Default: current working directory
- pattern (string): Expression to match (ls-like format)
(Accepts list of strings)
@Default: *
- avoid_empty_files (bool) : Ignore files with size 0
@Default: True
- rootless (bool) : When False, the name of the files are returned with
absolute path. Otherwise the root is removed.
@Default: False
@Yields file names
"""
if os.path.isfile(root): yield root
search_patterns = []
if not isinstance(pattern, list):
search_patterns.append(pattern)
else:
search_patterns = pattern
for pat in search_patterns:
for path, dirs, files in os.walk(os.path.abspath(root)):
for filename in fnmatch.filter(files, pat):
if not avoid_empty_files or os.path.getsize(
os.path.join(path, filename)) > 0:
SBIglobals.alert(
'debug', Path(), 'Found file {0}'.format(
os.path.join(path, filename)))
if not rootless:
yield os.path.join(path, filename)
else:
root = os.path.abspath(root) + "/"
yield os.path.join(path,
filename).replace(root, '')
def _build(self):
for dist_type in self.available_distance_types:
self._distance.setdefault(dist_type, None)
self._distance[self._threshold_type] = self.aminoacid.distance(self.nucleotide, dist_type = self._threshold_type)
SBIglobals.alert('deepdebug', self, '\tEvaluating distance {0:.3f} of {1}'.format(self._distance[self._threshold_type][2], self._threshold_type))
if float(self._distance[self._threshold_type][2]) <= self._threshold_distance and \
float(self._distance[self._threshold_type][2]) >= 0:
SBIglobals.alert('deepdebug', self, '\tDistance under threshold.')
self._underthreshold = True
for dist_type in self._distance:
if dist_type != self._threshold_type:
SBIglobals.alert('deepdebug', self, '\tGathering {0} distance'.format(dist_type))
self._distance[dist_type] = self.aminoacid.distance(self.nucleotide, dist_type = dist_type)
SBIglobals.alert('deepdebug', self, '\t\tDistance {0:.3f} of {1}'.format(self._distance[dist_type][2], dist_type))
def reduce(self, new_fasta_file, list_file, force=None):
'''
Reduces the {Fasta} by removing identical sequences.
@param: new_fasta_file
@pdef: name of the new fasta file
@ptype: {String}
@param: list_file
@pdef: name of the repetition list file
@ptype: {String}
@param: force
@pdef: overwrite previous files with the same name
@pdefault: _SBIglobals.overwrite_
@ptype: {Boolean}
@return: {Fasta} and {File} with the list of identical sequences.
'''
seq_md5 = {}
sequences = []
for seq in self.live_show():
md5 = seq.md5
if not md5 in seq_md5:
sequences.append(seq)
seq_md5.setdefault(md5, [])
else:
SBIg.alert(
'debug', self,
'{0} repeats of {1}'.format(seq.id, seq_md5[md5][0]))
seq_md5[md5].append(seq.id)
fasta = Fasta.build_multifasta(new_fasta_file, sequences, force)
listfile = File(list_file, 'w')
for md5 in seq_md5:
listfile.write('\t'.join(seq_md5[md5]) + '\n')
listfile.close()
return fasta, listfile
def write(self, output_file=None, format='PDB', force=None, clean=False):
"""
Writes the object in a specific format
@type output_file: String
@param output_file: File to write
@type format: String
@param format: Format of the file to print
"""
outfile = File(
file_name=output_file, action='w', overwrite=SBIg.decide_overwrite(force))
if format == 'PDB':
self._write_PDB_file(pdb_file=outfile, clean=clean)