chython 2.19__cp313-cp313-win_amd64.whl

chython/algorithms/standardize/reaction.py

@@ -0,0 +1,384 @@
+# -*- coding: utf-8 -*-
+#
+#  Copyright 2018-2026 Ramil Nugmanov <nougmanoff@protonmail.com>
+#  Copyright 2021 Timur Gimadiev <timur.gimadiev@gmail.com>
+#  Copyright 2024 Philippe Gantzer <p.gantzer@icredd.hokudai.ac.jp>
+#  This file is part of chython.
+#
+#  chython is free software; you can redistribute it and/or modify
+#  it under the terms of the GNU Lesser General Public License as published by
+#  the Free Software Foundation; either version 3 of the License, or
+#  (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+#  GNU Lesser General Public License for more details.
+#
+#  You should have received a copy of the GNU Lesser General Public License
+#  along with this program; if not, see <https://www.gnu.org/licenses/>.
+#
+from collections import defaultdict
+from ._reagents import *
+from ...exceptions import MappingError
+
+
+class StandardizeReaction:
+    __slots__ = ()
+
+    def canonicalize(self, *, fix_mapping: bool = True, logging=False, fix_tautomers=True) -> bool | list:
+        """
+        Convert molecules to canonical forms of functional groups and aromatic rings without explicit hydrogens.
+        Return True if in any molecule found not canonical group.
+
+        :param fix_mapping: Search AAM errors of functional groups.
+        :param logging: return processing log.
+        :param fix_tautomers: convert tautomers to canonical forms.
+        """
+        total = []
+        for n, m in enumerate(self.molecules()):
+            total.extend((n, x) for x in m.canonicalize(logging=True, fix_tautomers=fix_tautomers))
+
+        if fix_mapping:
+            total.extend(self.fix_mapping(logging=True))
+
+        if total:
+            self.flush_cache(keep_molecule_cache=True)
+        if logging:
+            return total
+        return bool(total)
+
+    def thiele(self, *, fix_tautomers=True) -> bool:
+        """
+        Convert structures to aromatic form.
+        Return True if in any molecule found kekule ring
+
+        :param fix_tautomers: convert tautomers to canonical forms.
+        """
+        total = False
+        for m in self.molecules():
+            if m.thiele(fix_tautomers=fix_tautomers):
+                total = True
+        if total:
+            self.flush_cache(keep_molecule_cache=True)
+        return total
+
+    def kekule(self, *, buffer_size=7, ignore_pyrrole_hydrogen=False) -> bool:
+        """
+        Convert structures to a kekule form.
+        Return True if in any molecule found aromatic ring
+
+        :param buffer_size: number of attempts of pyridine form searching.
+        :param ignore_pyrrole_hydrogen: ignore hydrogen on pyrrole to fix invalid rings like Cn1cc[nH]c1.
+        """
+        total = False
+        for m in self.molecules():
+            if m.kekule(buffer_size=buffer_size, ignore_pyrrole_hydrogen=ignore_pyrrole_hydrogen):
+                total = True
+        if total:
+            self.flush_cache(keep_molecule_cache=True)
+        return total
+
+    def clean_isotopes(self) -> bool:
+        """
+        Clean isotope marks for all molecules in reaction.
+        Returns True if in any molecule found isotope.
+        """
+        flag = False
+        for m in self.molecules():
+            if m.clean_isotopes():
+                flag = True
+        if flag:
+            self.flush_cache(keep_molecule_cache=True)
+        return flag
+
+    def clean_stereo(self):
+        """
+        Remove stereo data
+        """
+        for m in self.molecules():
+            m.clean_stereo()
+        self.flush_cache(keep_molecule_cache=True)
+
+    def implicify_hydrogens(self) -> int:
+        """
+        Remove explicit hydrogens if possible.
+
+        :return: number of removed hydrogens.
+        """
+        total = 0
+        for m in self.molecules():
+            total += m.implicify_hydrogens()
+        if total:
+            self.flush_cache(keep_molecule_cache=True)
+        return total
+
+    def explicify_hydrogens(self) -> int:
+        """
+        Add explicit hydrogens to atoms
+
+        :return: number of added atoms
+        """
+        total = 0
+        start_map = 0
+        for m in self.molecules():
+            map_ = max(m, default=0)
+            if map_ > start_map:
+                start_map = map_
+
+        mapping = defaultdict(list)
+        for m in self.reactants:
+            maps = m.explicify_hydrogens(_return_map=True, start_map=start_map + 1)
+            if maps:
+                for n, h in maps:
+                    mapping[n].append(h)
+                start_map = maps[-1][1]
+                total += len(maps)
+
+        for m in self.reagents:
+            maps = m.explicify_hydrogens(_return_map=True, start_map=start_map + 1)
+            if maps:
+                start_map = maps[-1][1]
+                total += len(maps)
+
+        for m in self.products:
+            maps = m.explicify_hydrogens(_return_map=True, start_map=start_map + 1)
+            if maps:
+                total += len(maps)
+                remap = {}
+                free = []
+                for n, h in maps:
+                    if n in mapping and mapping[n]:
+                        remap[h] = mapping[n].pop()
+                        free.append(h)
+                    elif free:
+                        remap[h] = start_map = free.pop(0)
+                    else:
+                        start_map = h
+                m.remap(remap)
+
+        if total:
+            self.flush_cache(keep_molecule_cache=True)
+        return total
+
+    def remove_reagents(self, *, keep_reagents: bool = False, mapping: bool = True) -> bool:
+        """
+        Place molecules, except reactants, to reagents list. Reagents - molecules which atoms not presented in products.
+        Mapping based approach remove molecules without reaction center.
+        Rule based approach remove equal molecules in reactants and products, and predefined reactants.
+
+        :param mapping: use atom-to-atom mapping to detect reagents, otherwise use predefined list of common reagents.
+        :param keep_reagents: delete reagents if False
+
+        Return True if any reagent found.
+        """
+        if mapping:
+            return self.__remove_reagents_mapping(keep_reagents)
+        return self.__remove_reagents_rules(keep_reagents)
+
+    def __remove_reagents_rules(self, keep_reagents):
+        if not self.reactants or not self.products:  # there is no reaction
+            return False
+
+        reactants_st1 = []
+        products_st1 = []
+        reagents_st1 = set(self.reagents)
+
+        # find the same molecules in reactants and products
+        for m in self.reactants:
+            if m in self.products:
+                reagents_st1.add(m)
+            else:
+                reactants_st1.append(m)
+        for m in self.products:
+            if m in self.reactants:
+                reagents_st1.add(m)
+            else:
+                products_st1.append(m)
+        if not reactants_st1 or not products_st1:
+            return False  # keep bad reaction as is
+
+        reactants_st2 = []
+        products_st2 = []
+        reagents_st2 = reagents_st1.copy()
+
+        # filter out predefined reagents
+        for m in reactants_st1:
+            if m in reagents_set:
+                reagents_st2.add(m)
+            else:
+                reactants_st2.append(m)
+        for m in products_st1:
+            if m in reagents_set:
+                reagents_st2.add(m)
+            else:
+                products_st2.append(m)
+        if not reactants_st2 or not products_st2:  # reaction contains only simple molecules. roll-back to step 1
+            reactants_st2 = reactants_st1
+            products_st2 = products_st1
+            reagents_st2 = reagents_st1
+
+        # remove reagents from reactants
+        tmp = []
+        for m in self.reagents:
+            tmp.append(m)
+            if m in reagents_st2:
+                reagents_st2.discard(m)
+        tmp.extend(reagents_st2)
+        reagents = tuple(tmp) if keep_reagents else ()
+
+        self._reactants = tuple(reactants_st2)
+        self._products = tuple(products_st2)
+        self._reagents = reagents
+        self.fix_positions()
+        return True
+
+    def __remove_reagents_mapping(self, keep_reagents):
+        cgr = ~self
+        if cgr.center_atoms:
+            active = set(cgr.center_atoms)
+            reactants = []
+            products = []
+            reagents = set(self.reagents)
+            for i in self.reactants:
+                if not active.isdisjoint(i):
+                    reactants.append(i)
+                else:
+                    reagents.add(i)
+            for i in self.products:
+                if not active.isdisjoint(i):
+                    products.append(i)
+                else:
+                    reagents.add(i)
+
+            # remove reagents from reactants
+            tmp = []
+            for m in self.reagents:
+                tmp.append(m)
+                if m in reagents:
+                    reagents.discard(m)
+            tmp.extend(reagents)
+            reagents = tuple(tmp) if keep_reagents else ()
+
+            if len(reactants) != len(self.reactants) or len(products) != len(self.products) or len(reagents) != len(self.reagents):
+                self._reactants = tuple(reactants)
+                self._products = tuple(products)
+                self._reagents = reagents
+                self.fix_positions()
+                return True
+            return False
+        raise MappingError("Reaction center is absent according to mapping")
+
+    def contract_ions(self) -> bool:
+        """
+        Contract ions into salts (Molecules with disconnected components).
+        Note: works only for unambiguous cases. e.g. equal anions/cations and different or equal cations/anions.
+
+        Return True if any ions contracted.
+        """
+        neutral, cations, anions, total = _sift_ions(self.reagents)
+        salts = _contract_ions(anions, cations, total)
+        if salts:
+            neutral.extend(salts)
+            self._reagents = tuple(neutral)
+            changed = True
+        else:
+            changed = False
+
+        neutral, cations, anions, total = _sift_ions(self.reactants)
+        salts = _contract_ions(anions, cations, total)
+        if salts:
+            anions_order = {frozenset(m): n for n, m in enumerate(anions)}
+            cations_order = {frozenset(m): n for n, m in enumerate(cations)}
+            neutral.extend(salts)
+            self._reactants = tuple(neutral)
+            changed = True
+        else:
+            anions_order = cations_order = {}
+
+        neutral, cations, anions, total = _sift_ions(self.products)
+        if cations and anions:
+            anions.sort(key=lambda x: anions_order.get(frozenset(x), -1))
+            cations.sort(key=lambda x: cations_order.get(frozenset(x), -1))
+        salts = _contract_ions(anions, cations, total)
+        if salts:
+            neutral.extend(salts)
+            self._products = tuple(neutral)
+            changed = True
+
+        if changed:
+            self.fix_positions()
+            return True
+        return False
+
+
+def _sift_ions(mols):
+    anions = []
+    cations = []
+    neutral = []
+    total = 0
+    for m in mols:
+        c = int(m)
+        total += c
+        if c > 0:
+            cations.append(m)
+        elif c < 0:
+            anions.append(m)
+        else:
+            neutral.append(m)
+    return neutral, cations, anions, total
+
+
+def _contract_ions(anions, cations, total):
+    if not anions or not cations:  # nothing to contract
+        return
+    # check ambiguous cases
+    if total > 0:
+        if len(cations) > 1:  # deficit of anions
+            # we have an excess of cations. we can't assign anions univocally
+            return  # unite is ambiguous
+        salt = cations[0]
+        shift_x = salt._fix_plane_mean(0) + 1
+        for x in anions:
+            shift_x = x._fix_plane_mean(shift_x) + 1
+            salt = salt | x
+        return [salt]
+    elif total < 0:
+        if len(anions) > 1:  # deficit of cations
+            # we have an excess of anions. we can't assign cations univocally
+            return  # unite is ambiguous
+        salt = anions[0]
+        shift_x = salt._fix_plane_mean(0) + 1
+        for x in cations:
+            shift_x = x._fix_plane_mean(shift_x) + 1
+            salt = salt | x
+        return [salt]
+    elif len(set(anions)) > 1 and len(set(cations)) > 1:  # different anions and cations
+        return
+
+    salts = []
+    anions = anions.copy()
+    cations = cations.copy()
+    while anions:
+        ct = cations.pop()
+        an = anions.pop()
+        shift_x = ct._fix_plane_mean(0) + 1
+        shift_x = an._fix_plane_mean(shift_x) + 1
+        salt = ct | an
+        while True:
+            c = int(salt)
+            if c > 0:
+                an = anions.pop()
+                shift_x = an._fix_plane_mean(shift_x) + 1
+                salt = salt | an
+            elif c < 0:
+                ct = cations.pop()
+                shift_x = ct._fix_plane_mean(shift_x) + 1
+                salt = salt | ct
+            else:
+                break
+        salts.append(salt)
+    return salts
+
+
+__all__ = ['StandardizeReaction']

chython/algorithms/standardize/resonance.py

@@ -0,0 +1,191 @@
+# -*- coding: utf-8 -*-
+#
+#  Copyright 2021-2026 Ramil Nugmanov <nougmanoff@protonmail.com>
+#  This file is part of chython.
+#
+#  chython is free software; you can redistribute it and/or modify
+#  it under the terms of the GNU Lesser General Public License as published by
+#  the Free Software Foundation; either version 3 of the License, or
+#  (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+#  GNU Lesser General Public License for more details.
+#
+#  You should have received a copy of the GNU Lesser General Public License
+#  along with this program; if not, see <https://www.gnu.org/licenses/>.
+#
+from ...exceptions import ValenceError
+
+
+# atomic number constants
+B = 5
+C = 6
+N = 7
+O = 8
+Si = 14
+P = 15
+S = 16
+As = 33
+Se = 34
+Te = 52
+
+
+class Resonance:
+    __slots__ = ()
+
+    def fix_resonance(self, *, logging=False) -> bool | list:
+        """
+        Transform biradical or dipole resonance structures into neutral form. Return True if structure form changed.
+
+        :param logging: return list of changed atoms.
+        """
+        atoms = self._atoms
+        bonds = self._bonds
+        entries, exits, rads, constrains, nitrogen_cat, nitrogen_ani, sulfur_cat = self.__entries()
+        hs = set()
+        while len(rads) > 1:
+            n = rads.pop()
+            for path in self.__find_delocalize_path(n, rads, constrains, True):
+                atoms[n]._is_radical = False
+                hs.add(n)
+                for n, m, b in path:
+                    hs.add(m)
+                    bonds[n][m]._order = b
+                atoms[m]._is_radical = False  # noqa
+                rads.discard(m)
+                break  # path found
+            # path not found. atom n keep as is
+        while entries and exits:
+            n = entries.pop()
+            for path in self.__find_delocalize_path(n, exits, constrains, False):
+                l, m, b = path[-1]
+                if n in nitrogen_cat and m in nitrogen_ani:
+                    continue
+
+                if m in sulfur_cat:  # prevent X-[S+]=X >> X=S=X
+                    if b != 1:
+                        continue
+                    atoms[m]._charge -= 1
+                else:  # check cations end valence.
+                    atoms[m]._charge -= 1  # reduce atom change and check valence
+                    try:
+                        atoms[m].valence_rules(sum(int(y) for x, y in bonds[m].items() if x != l) + b)
+                    except ValenceError:
+                        atoms[m]._charge += 1  # roll back
+                        continue
+
+                # succeed!
+                atoms[n]._charge += 1
+                hs.add(n)
+                for n, m, b in path:
+                    hs.add(m)
+                    bonds[n][m]._order = b
+                exits.discard(m)
+                break  # path from negative atom to positive atom found.
+            # path not found. keep negative atom n as is
+        if hs:
+            for n in hs:
+                self.calc_implicit(n)
+            self.flush_cache(keep_sssr=True, keep_components=True, keep_special_connectivity=True)
+            self.fix_stereo()
+            if logging:
+                return list(hs)
+            return True
+        elif logging:
+            return []
+        return False
+
+    def __find_delocalize_path(self, start, finish, constrains, odd_only):
+        bonds = self._bonds
+        stack = [(start, n, 0, b.order + 1) for n, b in bonds[start].items() if n in constrains and b.order < 3]
+        path = []
+        seen = {start}
+        while stack:
+            last, current, depth, order = stack.pop()
+            if len(path) > depth:
+                seen.difference_update(x for _, x, _ in path[depth:])
+                path = path[:depth]
+
+            path.append((last, current, order))
+
+            if current in finish:
+                if odd_only:  # radicals!
+                    if len(path) % 2:
+                        yield path
+                    else:  # invalid path
+                        continue
+                elif depth:  # one bonded ignored. we search double bond transfer! A=A-A >> A-A=A.
+                    yield path
+
+            depth += 1
+            seen.add(current)
+            diff = -1 if depth % 2 else 1
+            stack.extend((current, n, depth, bo) for n, b in bonds[current].items()
+                         if n not in seen and n in constrains and 1 <= (bo := b.order + diff) <= 3)
+
+    def __entries(self):
+        atoms = self._atoms
+        bonds = self._bonds
+        errors = {n for n, a in self.atoms() if a.implicit_hydrogens is None}
+
+        transfer = set()
+        entries = set()
+        exits = set()
+        rads = set()
+        nitrogen_cat = set()
+        nitrogen_ani = set()
+        sulfur_cat = set()
+        for n, a in self.atoms():
+            if a not in (B, C, N, O, Si, P, S, As, Se, Te):
+                # filter non-organic set, halogens and aromatics
+                continue
+            elif a.is_radical:
+                rads.add(n)
+            elif a.charge == -1:
+                if a == B and a.total_hydrogens:  # skip [BHx-]
+                    continue
+                elif (lb := len(bonds[n])) == 4 and a == B:  # skip [B-]X4
+                    continue
+                elif lb == 6 and a == P:  # skip [P-]X6
+                    continue
+                if n in errors:  # only valid anions accepted
+                    continue
+                entries.add(n)
+            elif a.charge == 1:
+                lb = len(bonds[n])
+                if a == N:
+                    if lb == 4:  # skip ammonia
+                        continue
+                    elif lb == 2 and a.hybridization == 3:  # skip Azide
+                        (n1, b1), (n2, b2) = bonds[n].items()
+                        an1 = atoms[n1]
+                        an2 = atoms[n2]
+                        if b1 == b2 == 2 and (an1.charge == -1 and an1 == N or an2.charge == -1 and an2 == N):
+                            continue
+                    elif lb == 3 and a.hybridization == 2:  # X=[N+](-X)-X - prevent N-N migration
+                        nitrogen_ani.add(n)
+                elif a == P and lb == 4:  # skip [P+]R4
+                    continue
+                elif a == S:
+                    if lb == 2 and a.hybridization == 2:  # ad-hoc for X-[S+]=X
+                        sulfur_cat.add(n)
+                    elif lb == 3 and a.hybridization == 1:  # ad-hoc for X-[S+](-X)-X
+                        continue
+                exits.add(n)
+            transfer.add(n)
+
+        if exits or entries:  # try to move cation to nitrogen. saturation fixup.
+            for n, a in self.atoms():
+                if a == N and not a.charge:
+                    if a.hybridization == 1 and a.neighbors <= 3:  # any amine - potential e-donor
+                        entries.add(n)
+                        nitrogen_cat.add(n)
+                    elif a.hybridization == 3 and a.neighbors == 1:  # N#X-[X-] >> [N-]=X=X
+                        exits.add(n)
+                        nitrogen_ani.add(n)
+        return entries, exits, rads, transfer, nitrogen_cat, nitrogen_ani, sulfur_cat
+
+
+__all__ = ['Resonance']

chython/algorithms/standardize/salts.py

@@ -0,0 +1,128 @@
+# -*- coding: utf-8 -*-
+#
+#  Copyright 2021-2026 Ramil Nugmanov <nougmanoff@protonmail.com>
+#  This file is part of chython.
+#
+#  chython is free software; you can redistribute it and/or modify
+#  it under the terms of the GNU Lesser General Public License as published by
+#  the Free Software Foundation; either version 3 of the License, or
+#  (at your option) any later version.
+#
+#  This program is distributed in the hope that it will be useful,
+#  but WITHOUT ANY WARRANTY; without even the implied warranty of
+#  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+#  GNU Lesser General Public License for more details.
+#
+#  You should have received a copy of the GNU Lesser General Public License
+#  along with this program; if not, see <https://www.gnu.org/licenses/>.
+#
+from ._salts import acids, rules
+
+
+# atomic number constants
+H = 1
+N = 7
+s_metals = {3, 4, 11, 12, 19, 20, 37, 38, 55, 56, 87, 88}
+s_metals_ammonia = {3, 4, 7, 11, 12, 19, 20, 37, 38, 55, 56, 87, 88}
+
+
+class Salts:
+    __slots__ = ()
+
+    def remove_metals(self, *, skip_elements: list[int] = None, logging=False) -> bool | list:
+        """
+        Remove disconnected S-metals and ammonia.
+
+        :param logging: return deleted atoms list.
+        :param skip_elements: skip elements from removing.
+        """
+        atoms = self._atoms
+        bonds = self._bonds
+        filters = s_metals_ammonia.difference(skip_elements) if skip_elements else s_metals_ammonia
+
+        metals = []
+        for n, a in atoms.items():
+            if not bonds[n] and a.atomic_number in filters:
+                metals.append(n)
+
+        if 0 < len(metals) < len(self):
+            for n in metals:
+                del atoms[n]
+                del bonds[n]
+
+            self.flush_cache(keep_sssr=True)
+            if logging:
+                return metals
+            return True
+        elif logging:
+            return []
+        return False
+
+    def remove_acids(self, *, logging=False) -> bool | list:
+        """
+        Remove common acids from organic bases salts.
+        Works only for neutral pairs like HA+B. Use `neutralize` before.
+
+        :param logging: return deleted atoms list.
+        """
+        if self.connected_components_count > 1:
+            log = []
+            for c in self.connected_components:
+                if self.substructure(c, recalculate_hydrogens=False) in acids:
+                    log.extend(c)
+            if 0 < len(log) < len(self):  # prevent singularity
+                atoms = self._atoms
+                bonds = self._bonds
+
+                for n in log:
+                    del atoms[n]
+                    del bonds[n]
+
+                self.flush_cache()
+                if logging:
+                    return log
+                return True
+        if logging:
+            return []
+        return False
+
+    def split_metal_salts(self, *, skip_elements: list[int] = None, logging=False) -> bool | list:
+        """
+        Split connected S-metal salts to cation/anion pairs.
+
+        :param logging: return deleted bonds list.
+        :param skip_elements: skip elements from splitting.
+        """
+        atoms = self._atoms
+        bonds = self._bonds
+        filters = s_metals.difference(skip_elements) if skip_elements else s_metals
+
+        metals = [n for n, a in atoms.items() if a.atomic_number in filters]
+        if metals:
+            acceptors = set()
+            log = []
+            for q in rules:
+                for mapping in q.get_mapping(self, automorphism_filter=False):
+                    acceptors.add(mapping[1])
+
+            for n in metals:
+                for m in acceptors & bonds[n].keys():
+                    if atoms[n].charge == 4:  # prevent overcharging
+                        break
+                    del bonds[n][m]
+                    del bonds[m][n]
+                    atoms[n]._charge += 1
+                    atoms[m]._charge -= 1
+                    log.append((n, m))
+            if log:
+                self.flush_cache()
+                self.fix_stereo()
+                if logging:
+                    return log
+                return True
+        if logging:
+            return []
+        return False
+
+
+__all__ = ['Salts']