예제 #1
0
    def getTransformation(self):
        """Returns the :class:`~.Transformation` used to superpose this
        conformation onto reference coordinates.  The transformation can
        be used to superpose original PDB file onto the reference PDB file."""

        if self._ensemble._trans is not None:
            return Transformation(self._ensemble._trans[self._index])
예제 #2
0
def buildBiomolecules(header, atoms, biomol=None):
    """Returns *atoms* after applying biomolecular transformations from *header*
    dictionary.  Biomolecular transformations are applied to all coordinate
    sets in the molecule.

    Some PDB files contain transformations for more than 1 biomolecules.  A
    specific set of transformations can be choosen using *biomol* argument.
    Transformation sets are identified by numbers, e.g. ``"1"``, ``"2"``, ...

    If multiple biomolecular transformations are provided in the *header*
    dictionary, biomolecules will be returned as
    :class:`.AtomGroup` instances in a :func:`list`.

    If the resulting biomolecule has more than 26 chains, the molecular
    assembly will be split into multiple :class:`.AtomGroup`
    instances each containing at most 26 chains.  These
    :class:`.AtomGroup` instances will be returned in a tuple.

    Note that atoms in biomolecules are ordered according to chain identifiers.
    """

    if not isinstance(header, dict):
        raise TypeError('header must be a dictionary')

    if not isinstance(atoms, Atomic):
        raise TypeError('atoms must be an Atomic instance')

    biomt = header.get('biomoltrans')
    if not isinstance(biomt, dict) or len(biomt) == 0:
        raise ValueError("header doesn't contain biomolecular transformations")

    if not isinstance(atoms, AtomGroup):
        atoms = atoms.copy()

    biomols = []
    if biomol is None:
        keys = list(biomt)
    else:
        biomol = str(biomol)
        if biomol in biomt:
            keys = [biomol]
        else:
            LOGGER.warn('Transformations for biomolecule {0} was not '
                        'found in the header dictionary.'.format(biomol))
            return None

    keys.sort()
    for i in keys:
        segnm = list('ABCDEFGHIJKLMNOPQRSTUVWXYZ' * 20)
        ags = []
        mt = biomt[i]
        # mt is a list, first item is list of chain identifiers
        # following items are lines corresponding to transformation
        # mt must have 3n + 1 lines
        if (len(mt)) % 4 != 0:
            LOGGER.warn('Biomolecular transformations {0} were not '
                        'applied'.format(i))
            continue

        for times in range(int((len(mt)) / 4)):
            rotation = np.zeros((3, 3))
            translation = np.zeros(3)
            line0 = np.fromstring(mt[times * 4 + 1], sep=' ')
            rotation[0, :] = line0[:3]
            translation[0] = line0[3]
            line1 = np.fromstring(mt[times * 4 + 2], sep=' ')
            rotation[1, :] = line1[:3]
            translation[1] = line1[3]
            line2 = np.fromstring(mt[times * 4 + 3], sep=' ')
            rotation[2, :] = line2[:3]
            translation[2] = line2[3]
            t = Transformation(rotation, translation)

            newag = atoms.select('chain ' + ' '.join(mt[0])).copy()
            if newag is None:
                continue
            newag.all.setSegnames(segnm.pop(0))
            for acsi in range(newag.numCoordsets()):
                newag.setACSIndex(acsi)
                newag = t.apply(newag)
            newag.setACSIndex(0)
            ags.append(newag)

        if ags:
            newag = ags.pop(0)
            while ags:
                newag += ags.pop(0)
            newag.setTitle('{0} biomolecule {1}'.format(atoms.getTitle(), i))
            biomols.append(newag)

    if biomols:
        if len(biomols) == 1:
            return biomols[0]
        else:
            return biomols
    else:
        return None
예제 #3
0
파일: header.py 프로젝트: prody/ProDy
def buildBiomolecules(header, atoms, biomol=None):
    """Returns *atoms* after applying biomolecular transformations from *header*
    dictionary.  Biomolecular transformations are applied to all coordinate
    sets in the molecule.

    Some PDB files contain transformations for more than 1 biomolecules.  A
    specific set of transformations can be choosen using *biomol* argument.
    Transformation sets are identified by numbers, e.g. ``"1"``, ``"2"``, ...

    If multiple biomolecular transformations are provided in the *header*
    dictionary, biomolecules will be returned as
    :class:`.AtomGroup` instances in a :func:`list`.

    If the resulting biomolecule has more than 26 chains, the molecular
    assembly will be split into multiple :class:`.AtomGroup`
    instances each containing at most 26 chains.  These
    :class:`.AtomGroup` instances will be returned in a tuple.

    Note that atoms in biomolecules are ordered according to chain identifiers.
    """

    if not isinstance(header, dict):
        raise TypeError('header must be a dictionary')

    if not isinstance(atoms, Atomic):
        raise TypeError('atoms must be an Atomic instance')

    biomt = header.get('biomoltrans')
    if not isinstance(biomt, dict) or len(biomt) == 0:
        LOGGER.warn("no biomolecular transformations found so original structure was used")
        return atoms

    if not isinstance(atoms, AtomGroup):
        atoms = atoms.copy()

    biomols = []
    if biomol is None:
        keys = list(biomt)
    else:
        biomol = str(biomol)
        if biomol in biomt:
            keys = [biomol]
        else:
            LOGGER.warn('Transformations for biomolecule {0} was not '
                        'found in the header dictionary.'.format(biomol))
            return None

    keys.sort()
    for i in keys:
        segnm = list('ABCDEFGHIJKLMNOPQRSTUVWXYZ'*20)
        ags = []
        mt = biomt[i]
        # mt is a list, first item is list of chain identifiers
        # following items are lines corresponding to transformation
        # mt must have 3n + 1 lines
        if (len(mt)) % 4 != 0:
            LOGGER.warn('Biomolecular transformations {0} were not '
                        'applied'.format(i))
            continue

        for times in range(int((len(mt)) / 4)):
            rotation = np.zeros((3, 3))
            translation = np.zeros(3)
            line0 = np.fromstring(mt[times*4+1], sep=' ')
            rotation[0, :] = line0[:3]
            translation[0] = line0[3]
            line1 = np.fromstring(mt[times*4+2], sep=' ')
            rotation[1, :] = line1[:3]
            translation[1] = line1[3]
            line2 = np.fromstring(mt[times*4+3], sep=' ')
            rotation[2, :] = line2[:3]
            translation[2] = line2[3]
            t = Transformation(rotation, translation)

            newag = atoms.select('chain ' + ' '.join(mt[0])).copy()
            if newag is None:
                continue
            newag.all.setSegnames(segnm.pop(0))
            for acsi in range(newag.numCoordsets()):
                newag.setACSIndex(acsi)
                newag = t.apply(newag)
            newag.setACSIndex(0)
            ags.append(newag)

        if ags:
            newag = ags.pop(0)
            while ags:
                newag += ags.pop(0)
            newag.setTitle('{0} biomolecule {1}'
                           .format(atoms.getTitle(), i))
            biomols.append(newag)
            
    if biomols:
        if len(biomols) == 1:
            return biomols[0]
        else:
            return biomols
    else:
        return None
예제 #4
0
파일: header.py 프로젝트: anindita85/ProDy
def buildBiomolecules(header, atoms, biomol=None):
    """Return *atoms* after applying biomolecular transformations from *header*
    dictionary.  Biomolecular transformations are applied to all coordinate 
    sets in the molecule.
    
    Some PDB files contain transformations for more than 1 biomolecules.  A 
    specific set of transformations can be choosen using *biomol* argument.
    Transformation sets are identified by numbers, e.g. ``"1"``, ``"2"``, ...
    
    If multiple biomolecular transformations are provided in the *header*
    dictionary, biomolecules will be returned as 
    :class:`~.AtomGroup` instances in a :func:`list`.  

    If the resulting biomolecule has more than 26 chains, the molecular 
    assembly will be split into multiple :class:`~.AtomGroup`
    instances each containing at most 26 chains.  These 
    :class:`~.AtomGroup` instances will be returned in a tuple.
   
    Note that atoms in biomolecules are ordered according to chain identifiers.
   
    """

    if not isinstance(header, dict):
        raise TypeError('header must be a dictionary')
    if not isinstance(atoms, Atomic):
        raise TypeError('atoms must be an Atomic instance')
    biomt = header.get('biomoltrans')
    if not isinstance(biomt, dict) or len(biomt) == 0:
        raise ValueError("header doesn't contain biomolecular transformations")
    
    if not isinstance(atoms, AtomGroup):
        atoms = atoms.copy()
    biomols = []
    if biomol is None: 
        keys = biomt.keys()
    else:
        biomol = str(biomol)
        if biomol in biomt:
            keys = [biomol]
        else:
            LOGGER.warning('Transformations for biomolecule {0:s} was not '
                           'found in the header dictionary.'.format(biomol))
            return None

    c_max = 26
    keys.sort()
    for i in keys: 
        chids = list('ABCDEFGHIJKLMNOPQRSTUVWXYZ'*20)
        ags = []
        mt = biomt[i]
        # mt is a list, first item is list of chain identifiers
        # following items are lines corresponding to transformation
        # mt must have 3n + 1 lines
        if (len(mt) - 1) % 3 != 0:
            LOGGER.warning('Biomolecular transformations {0:s} were not '
                           'applied'.format(i))
            continue
        chids_used = []
        for times in range((len(mt) - 1)/ 3):
            rotation = np.zeros((3,3))
            translation = np.zeros(3)
            line = np.fromstring(mt[times*3+1], sep=' ')
            rotation[0,:] = line[:3]
            translation[0] = line[3]
            line = np.fromstring(mt[times*3+2], sep=' ')
            rotation[1,:] = line[:3]
            translation[1] = line[3]
            line = np.fromstring(mt[times*3+3], sep=' ')
            rotation[2,:] = line[:3]
            translation[2] = line[3]
            t = Transformation(rotation, translation)
            
            for chid in mt[0]:
                chids_used.append(chid)
                newag = atoms.select('chain ' + chid).copy()
                if newag is None:
                    continue
                
                for acsi in range(newag.numCoordsets()):
                    newag.setACSIndex(acsi)
                    newag = t.apply(newag)
                newag.setACSIndex(0)
                ags.append(newag)
        if ags:
            # Handles the case when there is more atom groups than the number
            # of chain identifiers
            if len(chids_used) != len(set(chids_used)):
                for newag in ags:
                    newag.select('all').setChids(chids.pop(0))
            if len(ags) <= c_max:
                ags_ = [ags]
            else:
                ags_ = []
                for j in range(len(ags)/c_max + 1):
                    ags_.append(ags[j*c_max: (j+1)*c_max])
            parts = []
            for k, ags in enumerate(ags_):
                if not ags:
                    continue
                newag = ags.pop(0)
                while ags:
                    newag += ags.pop(0)
                if len(ags_) > 1:
                    newag.setTitle('{0:s} biomolecule {1:s} part {2:d}'
                                   .format(atoms.getTitle(), i, k+1))
                else:
                    newag.setTitle('{0:s} biomolecule {1:s}'
                                   .format(atoms.getTitle(), i))
                parts.append(newag)
            if len(parts) == 1:
                biomols.append(newag)
            else:
                biomols.append(tuple(parts))
    if biomols:
        if len(biomols) == 1:
            return biomols[0]
        else:
            return biomols
    else:
        return None