Update guesser docs (#4743)

* update docs * i've forgotten how sphinx works * replace ref with class * update changelog * fix whitespace * Update default_guesser.py Co-authored-by: Irfan Alibay <[email protected]> --------- Co-authored-by: Irfan Alibay <[email protected]>
MDAnalysis · Oct 20, 2024 · 101008b · 101008b
1 parent 9b69745
commit 101008b
Show file tree

Hide file tree

Showing 19 changed files with 252 additions and 35 deletions.
diff --git a/package/CHANGELOG b/package/CHANGELOG
@@ -23,6 +23,8 @@ The rules for this file:
  * 2.8.0
 
 Fixes
+ * Adds guessed attributes documentation back to each parser page
+   and updates overall guesser docs (Issue #4696)
  * Fix Bohrium (Bh) atomic mass in tables.py (PR #3753)
  * set `n_parts` to the total number of frames being analyzed if
    `n_parts` is bigger. (Issue #4685)

diff --git a/package/MDAnalysis/guesser/default_guesser.py b/package/MDAnalysis/guesser/default_guesser.py
@@ -27,9 +27,70 @@
 
 DefaultGuesser is a generic guesser class that has basic guessing methods.
 This class is a general purpose guesser that can be used with most topologies,
-but being generic makes it the less accurate among all guessers.
+but being generic makes it the least accurate among all guessers.
 
 
+Guessing behavior
+-----------------
+
+This section describes how each attribute is guessed by the DefaultGuesser.
+
+Masses
+~~~~~~
+
+We first attempt to look up the mass of an atom based on its element if the
+element TopologyAttr is available. If not, we attempt to lookup the mass based
+on the atom type (``type``) TopologyAttr. If neither of these is available, we
+attempt to guess the atom type based on the atom name (``name``) and then
+lookup the mass based on the guessed atom type.
+
+
+Types
+~~~~~
+
+We attempt to guess the atom type based on the atom name (``name``).
+The name is first stripped of any numbers and symbols, and then looked up in
+the :data:`MDAnalysis.guesser.tables.atomelements` table. If the name is not
+found, we continue checking variations of the name following the logic in
+:meth:`DefaultGuesser.guess_atom_element`. Ultimately, if no match is found,
+the first character of the stripped name is returned.
+
+Elements
+~~~~~~~~
+
+This follows the same method as guessing atom types.
+
+
+Bonds
+~~~~~
+
+Bonds are guessed based on the distance between atoms.
+See :meth:`DefaultGuesser.guess_bonds` for more details.
+
+Angles
+~~~~~~
+
+Angles are guessed based on the bonds between atoms.
+See :meth:`DefaultGuesser.guess_angles` for more details.
+
+Dihedrals
+~~~~~~~~~
+
+Dihedrals are guessed based on the angles between atoms.
+See :meth:`DefaultGuesser.guess_dihedrals` for more details.
+
+Improper Dihedrals
+~~~~~~~~~~~~~~~~~~
+
+Improper dihedrals are guessed based on the angles between atoms.
+See :meth:`DefaultGuesser.guess_improper_dihedrals` for more details.
+
+Aromaticities
+~~~~~~~~~~~~~
+
+Aromaticity is guessed using RDKit's GetIsAromatic method.
+See :meth:`DefaultGuesser.guess_aromaticities` for more details.
+
 
 
 
@@ -70,6 +131,23 @@ class DefaultGuesser(GuesserBase):
     You can use this guesser either directly through an instance, or through
     the :meth:`~MDAnalysis.core.universe.Universe.guess_TopologyAttrs` method.
 
+    Parameters
+    ----------
+    universe : Universe
+        The Universe to apply the guesser on
+    box : np.ndarray, optional
+        The box of the Universe. This is used for bond guessing.
+    vdwradii : dict, optional
+        Dict relating atom types: vdw radii. This is used for bond guessing
+    fudge_factor : float, optional
+        The factor by which atoms must overlap each other to be considered
+        a bond.  Larger values will increase the number of bonds found. [0.55]
+    lower_bound : float, optional
+        The minimum bond length. All bonds found shorter than this length
+        will be ignored. This is useful for parsing PDB with altloc records
+        where atoms with altloc A and B may be very close together and
+        there should be no chemical bond between them. [0.1]
+
     Examples
     --------
     to guess bonds for a universe::
@@ -84,8 +162,23 @@ class DefaultGuesser(GuesserBase):
     """
     context = 'default'
 
-    def __init__(self, universe, **kwargs):
-        super().__init__(universe, **kwargs)
+    def __init__(
+        self,
+        universe,
+        box=None,
+        vdwradii=None,
+        fudge_factor=0.55,
+        lower_bound=0.1,
+        **kwargs
+    ):
+        super().__init__(
+            universe,
+            box=box,
+            vdwradii=vdwradii,
+            fudge_factor=fudge_factor,
+            lower_bound=lower_bound,
+            **kwargs
+        )
         self._guesser_methods = {
             'masses': self.guess_masses,
             'types': self.guess_types,
@@ -212,8 +305,19 @@ def guess_types(self, atom_types=None, indices_to_guess=None):
     def guess_atom_element(self, atomname):
         """Guess the element of the atom from the name.
 
-        Looks in dict to see if element is found, otherwise it uses the first
-        character in the atomname. The table comes from CHARMM and AMBER atom
+        First all numbers and symbols are stripped from the name.
+        Then the name is looked up in the
+        :data:`MDAnalysis.guesser.tables.atomelements` table.
+        If the name is not found, we remove the last character or
+        first character from the name and check the table for both,
+        with a preference for removing the last character. If the name is
+        still not found, we iteratively continue to remove the last character
+        or first character until we find a match. If ultimately no match
+        is found, the first character of the stripped name is returned.
+        
+        If the input name is an empty string, an empty string is returned.
+
+        The table comes from CHARMM and AMBER atom
         types, where the first character is not sufficient to determine the
         atom type. Some GROMOS ions have also been added.
 
@@ -270,26 +374,11 @@ def guess_bonds(self, atoms=None, coords=None):
 
         Parameters
         ----------
-        atoms : AtomGroup
-             atoms for which bonds should be guessed
-        fudge_factor : float, optional
-            The factor by which atoms must overlap eachother to be considered a
-            bond. Larger values will increase the number of bonds found. [0.55]
-        vdwradii : dict, optional
-            To supply custom vdwradii for atoms in the algorithm. Must be a
-            dict of format {type:radii}. The default table of van der Waals
-            radii is hard-coded as :data:`MDAnalysis.guesser.tables.vdwradii`.
-            Any user defined vdwradii passed as an argument will supercede the
-            table values. [``None``]
-        lower_bound : float, optional
-            The minimum bond length. All bonds found shorter than this length
-            will be ignored. This is useful for parsing PDB with altloc records
-            where atoms with altloc A and B maybe very close together and
-            there should be no chemical bond between them. [0.1]
-        box : array_like, optional
-            Bonds are found using a distance search, if unit cell information
-            is given, periodic boundary conditions will be considered in the
-            distance search. [``None``]
+        atoms: AtomGroup
+            atoms for which bonds should be guessed
+        coords: np.ndarray, optional
+            coordinates of the atoms. If not provided, the coordinates
+            of the ``atoms`` in the universe are used.
 
         Returns
         -------

diff --git a/package/MDAnalysis/topology/CRDParser.py b/package/MDAnalysis/topology/CRDParser.py
@@ -33,6 +33,12 @@
 Residues are detected through a change is either resid or resname
 while segments are detected according to changes in segid.
 
+.. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
 .. _CRD: https://www.charmmtutorial.org/index.php/CHARMM:The_Basics
 
 
@@ -72,6 +78,13 @@ class CRDParser(TopologyReaderBase):
      - Resnums
      - Segids
 
+
+    .. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
     .. versionchanged:: 2.8.0
        Type and mass are not longer guessed here. Until 3.0 these will still be
        set by default through through universe.guess_TopologyAttrs() API.

diff --git a/package/MDAnalysis/topology/DLPolyParser.py b/package/MDAnalysis/topology/DLPolyParser.py
@@ -30,6 +30,12 @@
  - Atomnames
  - Atomids
 
+.. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
 .. _Poly: http://www.stfc.ac.uk/SCD/research/app/ccg/software/DL_POLY/44516.aspx
 
 Classes

diff --git a/package/MDAnalysis/topology/DMSParser.py b/package/MDAnalysis/topology/DMSParser.py
@@ -86,11 +86,17 @@ class DMSParser(TopologyReaderBase):
       Segment:
         - Segids
 
+    .. note::
+
+        By default, atomtypes will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
     .. _DESRES: http://www.deshawresearch.com
     .. _Desmond: http://www.deshawresearch.com/resources_desmond.html
     .. _DMS: http://www.deshawresearch.com/Desmond_Users_Guide-0.7.pdf
     .. versionchanged:: 2.8.0
-        Removed type and mass guessing (attributes guessing takes place now
+        Removed type guessing (attributes guessing takes place now
         through universe.guess_TopologyAttrs() API).
 
     """

diff --git a/package/MDAnalysis/topology/ExtendedPDBParser.py b/package/MDAnalysis/topology/ExtendedPDBParser.py
@@ -78,6 +78,12 @@ class ExtendedPDBParser(PDBParser.PDBParser):
      - bonds
      - formalcharges
 
+   .. note::
+
+      By default, atomtypes and masses will be guessed on Universe creation.
+      This may change in release 3.0.
+      See :ref:`Guessers` for more information.
+
 
     See Also
     --------

diff --git a/package/MDAnalysis/topology/FHIAIMSParser.py b/package/MDAnalysis/topology/FHIAIMSParser.py
@@ -66,6 +66,13 @@ class FHIAIMSParser(TopologyReaderBase):
     Creates the following attributes:
      - Atomnames
 
+
+    .. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
     .. versionchanged:: 2.8.0
         Removed type and mass guessing (attributes guessing takes place now
         through universe.guess_TopologyAttrs() API).

diff --git a/package/MDAnalysis/topology/GMSParser.py b/package/MDAnalysis/topology/GMSParser.py
@@ -73,6 +73,13 @@ class GMSParser(TopologyReaderBase):
      - names
      - atomic charges
 
+
+    .. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
     .. versionadded:: 0.9.1
     .. versionchanged:: 2.8.0
         Removed type and mass guessing (attributes guessing takes place now

diff --git a/package/MDAnalysis/topology/GROParser.py b/package/MDAnalysis/topology/GROParser.py
@@ -67,6 +67,12 @@ class GROParser(TopologyReaderBase):
       - atomids
       - atomnames
 
+    .. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
     .. versionchanged:: 2.8.0
         Removed type and mass guessing (attributes guessing takes place now
         through universe.guess_TopologyAttrs() API).

diff --git a/package/MDAnalysis/topology/ITPParser.py b/package/MDAnalysis/topology/ITPParser.py
@@ -473,6 +473,13 @@ class ITPParser(TopologyReaderBase):
     .. _ITP: http://manual.gromacs.org/current/reference-manual/topologies/topology-file-formats.html#molecule-itp-file
     .. _TOP: http://manual.gromacs.org/current/reference-manual/file-formats.html#top
 
+    .. note::
+
+        By default, atomtypes and masses will be guessed on Universe creation
+        if they are not read from the input file.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
     .. versionchanged:: 2.2.0
       no longer adds angles for water molecules with SETTLE constraint
     .. versionchanged:: 2.8.0

diff --git a/package/MDAnalysis/topology/LAMMPSParser.py b/package/MDAnalysis/topology/LAMMPSParser.py
@@ -27,6 +27,12 @@
 
 Parses data_ or dump_ files produced by LAMMPS_.
 
+.. note::
+
+    By default, masses will be guessed on Universe creation if they are not
+    read from the input file. This may change in release 3.0.
+    See :ref:`Guessers` for more information.
+
 .. _LAMMPS: http://lammps.sandia.gov/
 .. _data: DATA file format: :http://lammps.sandia.gov/doc/2001/data_format.html
 .. _dump: http://lammps.sandia.gov/doc/dump.html

diff --git a/package/MDAnalysis/topology/MMTFParser.py b/package/MDAnalysis/topology/MMTFParser.py
@@ -64,6 +64,12 @@
  - segid
  - model
 
+ .. note::
+
+    By default, masses will be guessed on Universe creation.
+    This may change in release 3.0.
+    See :ref:`Guessers` for more information.
+
 Classes
 -------
 

diff --git a/package/MDAnalysis/topology/MOL2Parser.py b/package/MDAnalysis/topology/MOL2Parser.py
@@ -78,6 +78,13 @@ class MOL2Parser(TopologyReaderBase):
      - Elements
 
 
+    .. note::
+
+        By default, masses will be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
+
     Notes
     -----
     Elements are obtained directly from the SYBYL atom types. If some atoms have

diff --git a/package/MDAnalysis/topology/PDBParser.py b/package/MDAnalysis/topology/PDBParser.py
@@ -35,6 +35,13 @@
 :mod:`~MDAnalysis.topology.ExtendedPDBParser`) that can handle residue
 numbers up to 99,999.
 
+.. note::
+
+    Atomtypes will be created from elements if they are present and valid.
+    Otherwise, they will be guessed on Universe creation.
+    By default, masses will also be guessed on Universe creation.
+    This may change in release 3.0.
+    See :ref:`Guessers` for more information.
 
 
 .. Note::

diff --git a/package/MDAnalysis/topology/PDBQTParser.py b/package/MDAnalysis/topology/PDBQTParser.py
@@ -32,6 +32,12 @@
 * Reads a PDBQT file line by line and does not require sequential atom numbering.
 * Multi-model PDBQT files are not supported.
 
+.. note::
+
+    By default, masses will be guessed on Universe creation.
+    This may change in release 3.0.
+    See :ref:`Guessers` for more information.
+
 Notes
 -----
 Only reads atoms and their names; connectivity is not

diff --git a/package/MDAnalysis/topology/PQRParser.py b/package/MDAnalysis/topology/PQRParser.py
@@ -80,6 +80,14 @@ class PQRParser(TopologyReaderBase):
      - Resnames
      - Segids
 
+     .. note::
+
+        Atomtypes will be read from the input file if they are present
+        (e.g. GROMACS PQR files). Otherwise, they will be guessed on Universe
+        creation. By default, masses will also be guessed on Universe creation.
+        This may change in release 3.0.
+        See :ref:`Guessers` for more information.
+
 
     .. versionchanged:: 0.9.0
        Read chainID from a PQR file and use it as segid (before we always used