Source code for MDAnalysis.topology.TOPParser

# -*- Mode: python; tab-width: 4; indent-tabs-mode:nil; coding:utf-8 -*-
# vim: tabstop=4 expandtab shiftwidth=4 softtabstop=4
#
# MDAnalysis --- https://www.mdanalysis.org
# Copyright (c) 2006-2017 The MDAnalysis Development Team and contributors
# (see the file AUTHORS for the full list of names)
#
# Released under the GNU Public Licence, v2 or any higher version
#
# Please cite your use of MDAnalysis in published work:
#
# R. J. Gowers, M. Linke, J. Barnoud, T. J. E. Reddy, M. N. Melo, S. L. Seyler,
# D. L. Dotson, J. Domanski, S. Buchoux, I. M. Kenney, and O. Beckstein.
# MDAnalysis: A Python package for the rapid analysis of molecular dynamics
# simulations. In S. Benthall and S. Rostrup editors, Proceedings of the 15th
# Python in Science Conference, pages 102-109, Austin, TX, 2016. SciPy.
# doi: 10.25080/majora-629e541a-00e
#
# N. Michaud-Agrawal, E. J. Denning, T. B. Woolf, and O. Beckstein.
# MDAnalysis: A Toolkit for the Analysis of Molecular Dynamics Simulations.
# J. Comput. Chem. 32 (2011), 2319--2327, doi:10.1002/jcc.21787
#

"""
AMBER PRMTOP topology parser
============================

Reads an AMBER top file to build the system.

Amber keywords are turned into the following attributes:

+----------------------------+----------------------+
| AMBER flag                 | MDAnalysis attribute |
+----------------------------+----------------------+
| ATOM_NAME                  | names                |
+----------------------------+----------------------+
| CHARGE                     | charges              |
+----------------------------+----------------------+
| ATOMIC_NUMBER              | elements             |
+----------------------------+----------------------+
| MASS                       | masses               |
+----------------------------+----------------------+
| BONDS_INC_HYDROGEN         | bonds                |
| BONDS_WITHOUT_HYDROGEN     |                      |
+----------------------------+----------------------+
| ANGLES_INC_HYDROGEN        | angles               |
| ANGLES_WITHOUT_HYDROGEN    |                      |
+----------------------------+----------------------+
| DIHEDRALS_INC_HYDROGEN     | dihedrals / improper |
| DIHEDRALS_WITHOUT_HYDROGEN |                      |
+----------------------------+----------------------+
| ATOM_TYPE_INDEX            | type_indices         |
+----------------------------+----------------------+
| AMBER_ATOM_TYPE            | types                |
+----------------------------+----------------------+
| RESIDUE_LABEL              | resnames             |
+----------------------------+----------------------+
| RESIDUE_POINTER            | residues             |
+----------------------------+----------------------+

TODO:
  Add support for Chamber-style topologies
  More stringent tests

.. Note::

   The Amber charge is converted to electron charges as used in
   MDAnalysis and other packages. To get back Amber charges, multiply
   by 18.2223.

   Chamber-style Amber topologies (i.e. topologies generated via parmed
   conversion of a CHARMM topology to an AMBER one) are not currently supported.
   Support will likely be added in future MDAnalysis releases.

.. _`PARM parameter/topology file specification`:
   http://ambermd.org/formats.html#topology

Classes
-------

.. autoclass:: TOPParser
   :members:
   :inherited-members:

"""
from __future__ import absolute_import, division

from six.moves import range, zip
from six import raise_from
import numpy as np
import functools
from math import ceil
import itertools

from . import guessers
from .tables import Z2SYMB
from ..lib.util import openany, FORTRANReader
from .base import TopologyReaderBase
from ..core.topology import Topology
from ..core.topologyattrs import (
    Atomnames,
    Atomtypes,
    Atomids,
    Charges,
    Elements,
    Masses,
    Resnames,
    Resids,
    Resnums,
    Segids,
    AtomAttr,
    Bonds,
    Angles,
    Dihedrals,
    Impropers
)

import warnings
import logging

logger = logging.getLogger('MDAnalysis.topology.TOPParser')


class TypeIndices(AtomAttr):
    """Numerical type of each Atom"""
    attrname = 'type_indices'
    singular = 'type_index'
    level = 'atom'


[docs]class TOPParser(TopologyReaderBase): """Reads topology information from an AMBER top file. Reads the following Attributes if in topology: - Atomnames - Charges - Masses - Elements - Atomtypes - Resnames - Type_indices - Bonds - Angles - Dihedrals (inc. impropers) Guesses the following attributes: - Elements (if not included in topology) The format is defined in `PARM parameter/topology file specification`_. The reader tries to detect if it is a newer (AMBER 12?) file format by looking for the flag "ATOMIC_NUMBER". .. _`PARM parameter/topology file specification`: http://ambermd.org/formats.html#topology .. versionchanged:: 0.7.6 parses both amber10 and amber12 formats .. versionchanged:: 0.19.0 parses bonds, angles, dihedrals, and impropers .. versionchanged:: 1.0.0 warns users that chamber-style topologies are not current supported """ format = ['TOP', 'PRMTOP', 'PARM7']
[docs] def parse(self, **kwargs): """Parse Amber PRMTOP topology file *filename*. Returns ------- A MDAnalysis Topology object """ # Sections that we grab as we parse the file sections = { "ATOM_NAME": (1, 20, self.parse_names, "name", 0), "CHARGE": (1, 5, self.parse_charges, "charge", 0), "ATOMIC_NUMBER": (1, 10, self.parse_elements, "elements", 0), "MASS": (1, 5, self.parse_masses, "mass", 0), "ATOM_TYPE_INDEX": (1, 10, self.parse_type_indices, "type_indices", 0), "AMBER_ATOM_TYPE": (1, 20, self.parse_types, "types", 0), "RESIDUE_LABEL": (1, 20, self.parse_resnames, "resname", 11), "RESIDUE_POINTER": (1, 10, self.parse_residx, "respoint", 11), "BONDS_INC_HYDROGEN": (3, 10, self.parse_bonded, "bondh", 2), "BONDS_WITHOUT_HYDROGEN": (3, 10, self.parse_bonded, "bonda", 3), "ANGLES_INC_HYDROGEN": (4, 10, self.parse_bonded, "angh", 4), "ANGLES_WITHOUT_HYDROGEN": (4, 10, self.parse_bonded, "anga", 5), "DIHEDRALS_INC_HYDROGEN": (5, 10, self.parse_bonded, "dihh", 6), "DIHEDRALS_WITHOUT_HYDROGEN": (5, 10, self.parse_bonded, "diha", 7) } attrs = {} # empty dict for attrs that we'll fill # Open and check top validity # Reading header info POINTERS with openany(self.filename) as self.topfile: header = next(self.topfile) if not header.startswith("%VE"): raise ValueError( "{0} is not a valid TOP file. %VE Missing in header" "".format(self.filename)) title = next(self.topfile).split() if not (title[1] == "TITLE"): # Raise a separate warning if Chamber-style TOP is detected if title[1] == "CTITLE": emsg = ("{0} is detected as a Chamber-style TOP file. " "At this time MDAnalysis does not support such " "topologies".format(self.filename)) else: emsg = ("{0} is not a valid TOP file. " "'TITLE' missing in header".format(self.filename)) raise ValueError(emsg) while not header.startswith('%FLAG POINTERS'): header = next(self.topfile) next(self.topfile) topremarks = [next(self.topfile).strip() for i in range(4)] sys_info = [int(k) for i in topremarks for k in i.split()] header = next(self.topfile) # grab the next section title next_section = header.split("%FLAG")[1].strip() while next_section is not None: try: (num_per_record, per_line, func, name, sect_num) = sections[next_section] except KeyError: def next_getter(): return self.skipper() else: num = sys_info[sect_num] * num_per_record numlines = (num // per_line) if num % per_line != 0: numlines += 1 attrs[name] = func(num_per_record, numlines) def next_getter(): return next(self.topfile) try: line = next_getter() # Capture case where section is empty w/ 1 empty line if numlines == 0 and not line.strip(): line = next_getter() except StopIteration: next_section = None else: try: next_section = line.split("%FLAG")[1].strip() except IndexError: raise_from( IndexError(( "%FLAG section not found, formatting error " "for PARM7 file {0} ").format(self.filename)), None) # strip out a few values to play with them n_atoms = len(attrs['name']) resptrs = attrs.pop('respoint') resptrs.append(n_atoms) residx = np.zeros(n_atoms, dtype=np.int32) for i, (x, y) in enumerate(zip(resptrs[:-1], resptrs[1:])): residx[x:y] = i n_res = len(attrs['resname']) # Deal with recreating bonds and angle records here attrs['bonds'] = Bonds([i for i in itertools.chain( attrs.pop('bonda'), attrs.pop('bondh'))]) attrs['angles'] = Angles([i for i in itertools.chain( attrs.pop('anga'), attrs.pop('angh'))]) attrs['dihedrals'], attrs['impropers'] = self.parse_dihedrals( attrs.pop('diha'), attrs.pop('dihh')) # Guess elements if not in topology or if any dummy atoms are found if not 'elements' in attrs: msg = ("ATOMIC_NUMBER record not found, guessing atom elements " "based on their atom types") logger.warning(msg) warnings.warn(msg) attrs['elements'] = Elements( guessers.guess_types(attrs['types'].values), guessed=True) elif np.any(attrs['elements'].values == "DUMMY"): # This approach assumes np.any is much faster than np.where attrs['elements']._guessed = True for dummy_idx in np.where(attrs['elements'].values == 'DUMMY')[0]: atom_type = attrs['types'].values[dummy_idx] atom_ele = guessers.guess_atom_element(atom_type) msg = ("Unknown ATOMIC_NUMBER value found, guessing atom " "element from type: {0} assigned to {1}".format( atom_type, atom_ele)) logger.warning(msg) warnings.warn(msg) attrs['elements'].values[dummy_idx] = atom_ele # atom ids are mandatory attrs['atomids'] = Atomids(np.arange(n_atoms) + 1) attrs['resids'] = Resids(np.arange(n_res) + 1) attrs['resnums'] = Resnums(np.arange(n_res) + 1) attrs['segids'] = Segids(np.array(['SYSTEM'], dtype=object)) top = Topology(n_atoms, n_res, 1, attrs=list(attrs.values()), atom_resindex=residx, residue_segindex=None) return top
[docs] def skipper(self): """TOPParser :class: helper function, skips lines of input parm7 file until we find the next %FLAG entry and return that Returns ------- line : string String containing the current line of the parm7 file """ line = next(self.topfile) while not line.startswith("%FLAG"): line = next(self.topfile) return line
[docs] def parse_names(self, num_per_record, numlines): """Extracts atoms names from parm7 file Parameters ---------- num_per_record : int The number of entries for each record in the section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- attr : :class:`Atomnames` A :class:`Atomnames` instance containing the names of each atom as defined in the parm7 file """ vals = self.parsesection_mapper(numlines, lambda x: x) attr = Atomnames(np.array(vals, dtype=object)) return attr
[docs] def parse_resnames(self, num_per_record, numlines): """Extracts the names of each residue Parameters ---------- num_per_record : int The number of entries for each recrod in section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- attr : :class:`Resnames` A :class:`Resnames` instance containing the names of each residue as defined in the parm7 file """ vals = self.parsesection_mapper(numlines, lambda x: x) attr = Resnames(np.array(vals, dtype=object)) return attr
[docs] def parse_charges(self, num_per_record, numlines): """Extracts the partial charges for each atom Parameters ---------- num_per_record : int The number of entries for each record in section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- attr : :class:`Charges` A :class:`Charges` instance containing the partial charges of each atom as defined in the parm7 file """ vals = self.parsesection_mapper(numlines, lambda x: float(x)) charges = np.array(vals, dtype=np.float32) charges /= 18.2223 # to electron charge units attr = Charges(charges) return attr
[docs] def parse_masses(self, num_per_record, numlines): """Extracts the mass of each atom Parameters ---------- num_per_record : int The number of entries for each record in section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- attr : :class:`Masses` A :class:`Masses` instance containing the mass of each atom as defined in the parm7 file """ vals = self.parsesection_mapper(numlines, lambda x: float(x)) attr = Masses(vals) return attr
[docs] def parse_elements(self, num_per_record, numlines): """Extracts the atomic numbers of each atom and converts to element type Parameters ---------- num_per_record : int The number of entries for each record in section(unused input) numlines : int The number of lines to be pasred in current section Returns ------- attr : :class:`Elements` A :class:`Elements` instance containing the element of each atom as defined in the parm7 file Note ---- If the record contains atomic numbers <= 0, these will be treated as dummy elements and an attempt will be made to guess the element based on atom type. See issue #2306 for more details. """ vals = self.parsesection_mapper( numlines, lambda x: Z2SYMB[int(x)] if int(x) > 0 else "DUMMY") attr = Elements(np.array(vals, dtype=object)) return attr
[docs] def parse_types(self, num_per_record, numlines): """Extracts the force field atom types of each atom Parameters ---------- num_per_record : int The number of entries for each record in section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- attr : :class:`Atomtypes` A :class:`Atomtypes` instance containing the atom types for each atom as defined in the parm7 file """ vals = self.parsesection_mapper(numlines, lambda x: x) attr = Atomtypes(np.array(vals, dtype=object)) return attr
[docs] def parse_type_indices(self, num_per_record, numlines): """Extracts the index of atom types of the each atom involved in Lennard Jones (6-12) interactions. Parameters ---------- num_per_record : int The number of entries for each record in section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- attr :class:`TypeIndices` A :class:`TypeIndices` instance containing the LJ 6-12 atom type index for each atom """ vals = self.parsesection_mapper(numlines, lambda x: int(x)) attr = TypeIndices(np.array(vals, dtype=np.int32)) return attr
[docs] def parse_residx(self, num_per_record, numlines): """Extracts the residue pointers for each atom Parameters ---------- num_per_record : int The number of entries for each record in section (unused input) numlines : int The number of lines to be parsed in current section Returns ------- vals : list of int A list of zero-formatted residue pointers for each atom """ vals = self.parsesection_mapper(numlines, lambda x: int(x) - 1) return vals
[docs] def parse_chunks(self, data, chunksize): """Helper function to parse AMBER PRMTOP bonds/angles. Parameters ---------- data : list of int Input list of the parm7 bond/angle section, zero-indexed num_per_record : int The number of entries for each record in the input list Returns ------- vals : list of int tuples A list of tuples containing the atoms involved in a given bonded interaction Note ---- In the parm7 format this information is structured in the following format: [ atoms 1:n, internal index ] Where 1:n represent the ids of the n atoms involved in the bond/angle and the internal index links to a given set of FF parameters. Therefore, to extract the required information, we split out the list into chunks of size num_per_record, and only extract the atom ids. """ vals = [tuple(data[x:x+chunksize-1]) for x in range(0, len(data), chunksize)] return vals
[docs] def parse_bonded(self, num_per_record, numlines): """Extracts bond information from PARM7 format files Parameters ---------- num_per_record : int The number of entries for each record in section numlines : int The number of lines to be parsed for this section Note ---- For the bond/angle sections of parm7 files, the atom numbers are set to coordinate array index values. As detailed in http://ambermd.org/formats.html to recover the actual atom number, one should divide the values by 3 and add 1. Here, since we want to satisfy zero-indexing, we only divide by 3. """ fields = self.parsesection_mapper(numlines, lambda x: int(x) // 3) section = self.parse_chunks(fields, num_per_record) return section
[docs] def parsesection_mapper(self, numlines, mapper): """Parses FORTRAN formatted section, and returns a list of all entries in each line Parameters ---------- numlines : int The number of lines to be parsed in this section mapper : lambda operator Operator to format entries in current section Returns ------- section : list A list of all entries in a given parm7 section """ section = [] y = next(self.topfile).strip("%FORMAT(") y.strip(")") x = FORTRANReader(y) for i in range(numlines): l = next(self.topfile) for j in range(len(x.entries)): val = l[x.entries[j].start:x.entries[j].stop].strip() if val: section.append(mapper(val)) return section
[docs] def parse_dihedrals(self, diha, dihh): """Combines hydrogen and non-hydrogen containing AMBER dihedral lists and extracts sublists for conventional dihedrals and improper angles Parameters ---------- diha : list of tuples The atom ids of dihedrals not involving hydrogens dihh : list of tuples The atom ids of dihedrals involving hydrogens Returns ------- dihedrals : :class:`Dihedrals` A :class:`Dihedrals` instance containing a list of all unique dihedrals as defined by the parm7 file impropers : :class:`Impropers` A :class:`Impropers` instance containing a list of all unique improper dihedrals as defined by the parm7 file Note ---- As detailed in http://ambermd.org/formats.html, the dihedral sections of parm7 files contain information about both conventional dihedrals and impropers. The following must be accounted for: 1) If the fourth atom in a dihedral entry is given a negative value, this indicates that it is an improper. 2) If the third atom in a dihedral entry is given a negative value, this indicates that it 1-4 NB interactions are ignored for this dihedrals. This could be due to the dihedral within a ring, or if it is part of a multi-term dihedral definition or if it is an improper. """ improp = [] dihed = [] for i in itertools.chain(diha, dihh): if i[3] < 0: improp.append(i[:2]+(abs(i[2]),)+(abs(i[3]),)) elif i[2] < 0: vals = i[:2] + (abs(i[2]),) + i[3:] dihed.append(vals) else: dihed.append(i) dihed = sorted(set(dihed)) dihedrals = Dihedrals(dihed) impropers = Impropers(improp) return dihedrals, impropers