RNApolis 0.4.3__tar.gz → 0.4.6__tar.gz

{rnapolis-0.4.3/src/RNApolis.egg-info → rnapolis-0.4.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: RNApolis
-Version: 0.4.3
+Version: 0.4.6
 Summary: A Python library containing RNA-related bioinformatics functions and classes
 Home-page: https://github.com/tzok/rnapolis-py
 Author: Tomasz Zok

{rnapolis-0.4.3 → rnapolis-0.4.6}/setup.py

@@ -5,7 +5,7 @@ with open("README.md") as f:
 
 setup(
     name="RNApolis",
-    version="0.4.3",
+    version="0.4.6",
     packages=["rnapolis"],
     package_dir={"": "src"},
     author="Tomasz Zok",

{rnapolis-0.4.3 → rnapolis-0.4.6/src/RNApolis.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: RNApolis
-Version: 0.4.3
+Version: 0.4.6
 Summary: A Python library containing RNA-related bioinformatics functions and classes
 Home-page: https://github.com/tzok/rnapolis-py
 Author: Tomasz Zok

{rnapolis-0.4.3 → rnapolis-0.4.6}/src/rnapolis/parser.py

@@ -1,7 +1,10 @@
 import logging
 from typing import IO, Dict, List, Optional, Tuple, Union
 
+import numpy as np
 from mmcif.io.IoAdapterPy import IoAdapterPy
+from scipy.spatial import KDTree
+
 from rnapolis.common import ResidueAuth, ResidueLabel
 from rnapolis.tertiary import BASE_ATOMS, Atom, Residue3D, Structure3D
 
@@ -53,10 +56,10 @@ def parse_cif(
 
     io_adapter = IoAdapterPy()
     data = io_adapter.readFile(cif.name)
-    atoms: List[Atom] = []
+    atoms_to_process: List[Atom] = []
     modified: Dict[Union[ResidueLabel, ResidueAuth], str] = {}
-    sequence_by_entity = {}
-    is_nucleic_acid_by_entity = {}
+    sequence_by_entity: Dict[str, str] = {}
+    is_nucleic_acid_by_entity: Dict[str, bool] = {}
 
     if data:
         atom_site = data[0].getObj("atom_site")
@@ -136,7 +139,7 @@ def parse_cif(
                     else None
                 )
 
-                atoms.append(
+                atoms_to_process.append(
                     Atom(
                         label_entity_id,
                         label,
@@ -216,6 +219,7 @@ def parse_cif(
                 if entity_id and pdbx_seq_one_letter_code_can:
                     sequence_by_entity[entity_id] = pdbx_seq_one_letter_code_can
 
+    atoms = filter_clashing_atoms(atoms_to_process)
     return atoms, modified, sequence_by_entity, is_nucleic_acid_by_entity
 
 
@@ -228,7 +232,7 @@ def parse_pdb(
     Dict[str, bool],
 ]:
     pdb.seek(0)
-    atoms: List[Atom] = []
+    atoms_to_process: List[Atom] = []
     modified: Dict[Union[ResidueLabel, ResidueAuth], str] = {}
     model = 1
 
@@ -236,9 +240,6 @@ def parse_pdb(
         if line.startswith("MODEL"):
             model = int(line[10:14].strip())
         elif line.startswith("ATOM") or line.startswith("HETATM"):
-            alternate_location = line[16]
-            if alternate_location != " ":
-                continue
             atom_name = line[12:16].strip()
             residue_name = line[17:20].strip()
             chain_identifier = line[21]
@@ -251,7 +252,10 @@ def parse_pdb(
             auth = ResidueAuth(
                 chain_identifier, residue_number, insertion_code, residue_name
             )
-            atoms.append(Atom(None, None, auth, model, atom_name, x, y, z, occupancy))
+
+            atoms_to_process.append(
+                Atom(None, None, auth, model, atom_name, x, y, z, occupancy)
+            )
         elif line.startswith("MODRES"):
             original_name = line[12:15]
             chain_identifier = line[16]
@@ -263,6 +267,7 @@ def parse_pdb(
             )
             modified[auth] = standard_residue_name
 
+    atoms = filter_clashing_atoms(atoms_to_process)
     return atoms, modified, {}, {}
 
 
@@ -392,3 +397,36 @@ def try_parse_int(s: str) -> Optional[int]:
         return int(s)
     except ValueError:
         return None
+
+
+def filter_clashing_atoms(atoms: List[Atom], clash_distance: float = 0.5) -> List[Atom]:
+    # First, remove duplicate atoms
+    unique_atoms = {}
+
+    for i, atom in enumerate(atoms):
+        key = (atom.label, atom.auth, atom.name)
+        if key not in unique_atoms or atom.occupancy > unique_atoms[key].occupancy:
+            unique_atoms[key] = atom
+
+    unique_atoms_list = list(unique_atoms.values())
+
+    # Now handle clashing atoms
+    coords = np.array([(atom.x, atom.y, atom.z) for atom in unique_atoms_list])
+    tree = KDTree(coords)
+
+    pairs = tree.query_pairs(r=clash_distance)
+
+    atoms_to_keep = set(range(len(unique_atoms_list)))
+
+    for i, j in pairs:
+        if (
+            unique_atoms_list[i].occupancy is None
+            or unique_atoms_list[j].occupancy is None
+        ):
+            continue
+        if unique_atoms_list[i].occupancy > unique_atoms_list[j].occupancy:
+            atoms_to_keep.discard(j)
+        else:
+            atoms_to_keep.discard(i)
+
+    return [unique_atoms_list[i] for i in atoms_to_keep]

{rnapolis-0.4.3 → rnapolis-0.4.6}/src/rnapolis/tertiary.py

@@ -7,6 +7,7 @@ from typing import Dict, List, Optional, Set, Tuple, Union
 
 import numpy
 import numpy.typing
+
 from rnapolis.common import (
     BasePair,
     BpSeq,
@@ -16,6 +17,7 @@ from rnapolis.common import (
     Residue,
     ResidueAuth,
     ResidueLabel,
+    Saenger,
     Stacking,
 )
 
@@ -521,13 +523,47 @@ class Mapping2D3D:
 
     @cached_property
     def bpseq(self) -> BpSeq:
-        return self.__generate_bpseq(
-            [
-                base_pair
-                for base_pair in self.base_pairs
-                if base_pair.is_canonical and base_pair.nt1 < base_pair.nt2
-            ]
-        )
+        def pair_scoring_function(pair: BasePair3D) -> int:
+            if pair.saenger is not None:
+                if pair.saenger in (Saenger.XIX, Saenger.XX):
+                    return 0, pair.nt1, pair.nt2
+                else:
+                    return 1, pair.nt1, pair.nt2
+
+            sequence = "".join(
+                sorted(
+                    [
+                        pair.nt1_3d.one_letter_name.upper(),
+                        pair.nt2_3d.one_letter_name.upper(),
+                    ]
+                )
+            )
+            if sequence in ("AU", "AT", "CG"):
+                return 0, pair.nt1, pair.nt2
+            return 1, pair.nt1, pair.nt2
+
+        canonical = [
+            base_pair
+            for base_pair in self.base_pairs
+            if base_pair.is_canonical and base_pair.nt1 < base_pair.nt2
+        ]
+
+        while True:
+            matches = defaultdict(set)
+
+            for base_pair in canonical:
+                matches[base_pair.nt1_3d].add(base_pair)
+                matches[base_pair.nt2_3d].add(base_pair)
+
+            for pairs in matches.values():
+                if len(pairs) > 1:
+                    pairs = sorted(pairs, key=pair_scoring_function)
+                    canonical.remove(pairs[-1])
+                    break
+            else:
+                break
+
+        return self.__generate_bpseq(canonical)
 
     def __generate_bpseq(self, base_pairs):
         result: Dict[int, List] = {}

{rnapolis-0.4.3 → rnapolis-0.4.6}/tests/test_annotator.py

@@ -43,3 +43,14 @@ def test_8btk():
     with open("tests/8btk_B7.cif") as f:
         structure3d = read_3d_structure(f, 1)
     assert extract_secondary_structure(structure3d, 1) is not None
+
+
+def test_488d():
+    """
+    There are clashing residues 151 in chains B and D. The clash is caused by occupancy factors less than 1.
+    """
+    with open("tests/488d.pdb") as f:
+        structure3d = read_3d_structure(f)
+
+    base_interactions = extract_base_interactions(structure3d)
+    assert base_interactions is not None

{rnapolis-0.4.3 → rnapolis-0.4.6}/tests/test_common.py

@@ -1,7 +1,9 @@
 from collections import Counter
 
+import orjson
 from hypothesis import given, settings
 from hypothesis import strategies as st
+
 from rnapolis.common import (
     BaseInteractions,
     BasePair,
@@ -10,13 +12,17 @@ from rnapolis.common import (
     BpSeq,
     DotBracket,
     Interaction,
+    LeontisWesthof,
     MultiStrandDotBracket,
     OtherInteraction,
     Residue,
     ResidueAuth,
     ResidueLabel,
+    Saenger,
     Stacking,
 )
+from rnapolis.parser import read_3d_structure
+from rnapolis.tertiary import Mapping2D3D
 
 
 @given(st.from_type(ResidueLabel))
@@ -136,3 +142,41 @@ def test_multi_strand_dot_bracket():
     assert dot_bracket.strands[1].structure == (
         ")))))))))))..(((...[[[[[[...)))......)))))...]]]]]][[[[[.((((((]]]]].....((((((......((((((....)))))).......))))))..))))))."
     )
+
+
+def test_conflicted_base_pairs():
+    with open("tests/1A1T_1_B-rnaview.json", "rb") as f:
+        data = orjson.loads(f.read())
+
+    base_pairs = []
+
+    for obj in data.get("basePairs", []):
+        nt1 = Residue(
+            None,
+            ResidueAuth(
+                obj["nt1"]["auth"]["chain"],
+                obj["nt1"]["auth"]["number"],
+                obj["nt1"]["auth"]["icode"],
+                obj["nt1"]["auth"]["name"],
+            ),
+        )
+        nt2 = Residue(
+            None,
+            ResidueAuth(
+                obj["nt2"]["auth"]["chain"],
+                obj["nt2"]["auth"]["number"],
+                obj["nt2"]["auth"]["icode"],
+                obj["nt2"]["auth"]["name"],
+            ),
+        )
+        lw = LeontisWesthof(obj["lw"])
+        saenger = Saenger(obj["saenger"]) if obj["saenger"] else None
+        base_pairs.append(BasePair(nt1, nt2, lw, saenger))
+
+    with open("tests/1A1T_1_B.cif") as f:
+        structure3d = read_3d_structure(f)
+
+    mapping = Mapping2D3D(structure3d, base_pairs, [], True)
+    assert (
+        mapping.dot_bracket == ">strand_B\nGGACUAGCGGAGGCUAGUCC\n((((((((....))))))))"
+    )

{rnapolis-0.4.3 → rnapolis-0.4.6}/tests/test_parser.py

@@ -16,3 +16,18 @@ def test_1ato():
         structure3d = read_3d_structure(f)
     sequence = "".join([residue.one_letter_name for residue in structure3d.residues])
     assert sequence == "GGCACCUCCUCGCGGUGCC"
+
+
+def test_4qln_no_duplicate_atoms():
+    for ext in (".pdb", ".cif"):
+        with open(f"tests/4qln{ext}") as f:
+            structure3d = read_3d_structure(f)
+
+        chain_a = [r for r in structure3d.residues if r.auth.chain == "A"]
+        residues_to_check = [r for r in chain_a if r.auth.number in (18, 19, 20)]
+
+        for residue in residues_to_check:
+            atom_names = [atom.name for atom in residue.atoms]
+            assert len(atom_names) == len(
+                set(atom_names)
+            ), f"Duplicate atoms found in residue {residue.auth}"