RNApolis 0.4.2__tar.gz → 0.4.4__tar.gz

{rnapolis-0.4.2/src/RNApolis.egg-info → rnapolis-0.4.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: RNApolis
-Version: 0.4.2
+Version: 0.4.4
 Summary: A Python library containing RNA-related bioinformatics functions and classes
 Home-page: https://github.com/tzok/rnapolis-py
 Author: Tomasz Zok

{rnapolis-0.4.2 → rnapolis-0.4.4}/setup.py

@@ -5,7 +5,7 @@ with open("README.md") as f:
 
 setup(
     name="RNApolis",
-    version="0.4.2",
+    version="0.4.4",
     packages=["rnapolis"],
     package_dir={"": "src"},
     author="Tomasz Zok",

{rnapolis-0.4.2 → rnapolis-0.4.4/src/RNApolis.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: RNApolis
-Version: 0.4.2
+Version: 0.4.4
 Summary: A Python library containing RNA-related bioinformatics functions and classes
 Home-page: https://github.com/tzok/rnapolis-py
 Author: Tomasz Zok

{rnapolis-0.4.2 → rnapolis-0.4.4}/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,
 )
 
@@ -497,34 +499,71 @@ class Mapping2D3D:
 
     @cached_property
     def strands_sequences(self) -> List[Tuple[str, str]]:
-        result = defaultdict(list)
+        nucleotides = list(filter(lambda r: r.is_nucleotide, self.structure3d.residues))
 
-        for i, residue in enumerate(
-            filter(lambda r: r.is_nucleotide, self.structure3d.residues)
-        ):
-            if i > 0 and self.find_gaps:
-                previous = self.structure3d.residues[i - 1]
+        if not nucleotides:
+            return []
 
-                if (
-                    not previous.is_connected(residue)
-                    and previous.chain == residue.chain
-                ):
-                    for k in range(residue.number - previous.number - 1):
-                        result[residue.chain].append("?")
+        result = [(nucleotides[0].chain, [nucleotides[0].one_letter_name])]
 
-            result[residue.chain].append(residue.one_letter_name)
+        for i in range(1, len(nucleotides)):
+            previous = nucleotides[i - 1]
+            residue = nucleotides[i]
 
-        return [(chain, "".join(sequence)) for chain, sequence in result.items()]
+            if residue.chain != previous.chain:
+                result.append((residue.chain, [residue.one_letter_name]))
+            else:
+                if self.find_gaps:
+                    if not previous.is_connected(residue):
+                        for k in range(residue.number - previous.number - 1):
+                            result[-1][1].append("?")
+                result[-1][1].append(residue.one_letter_name)
+
+        return [(chain, "".join(sequence)) for chain, sequence in result]
 
     @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-0.4.2 → rnapolis-0.4.4}/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((((((((....))))))))"
+    )