gemmi-protools 1.0.0__py3-none-any.whl → 1.0.2__py3-none-any.whl

gemmi_protools/io/reader.py

@@ -1,3 +1,5 @@
+import gzip
+import io
 import itertools
 import pathlib
 import random
@@ -274,19 +276,7 @@ class StructureParser(object):
         else:
             raise ValueError("structure must be gemmi.Structure or None")
 
-        self.STRUCT.setup_entities()
-        self.STRUCT.assign_serial_numbers()
-
-        self.STRUCT.renumber_models()
-        if len(self.STRUCT) > 1:
-            for idx in range(1, len(self.STRUCT)):
-                del self.STRUCT[idx]
-
-        self.MODEL = self.STRUCT[0]
-        self.STRUCT.remove_alternative_conformations()
-        self.STRUCT.remove_hydrogens()
-        self.STRUCT.remove_empty_chains()
-        self._update_full_sequences()
+        self._init_struct()
 
         info_map = dict(self.STRUCT.info)
         pdb_code = info_map.get("_entry.id", "").lower()
@@ -300,6 +290,22 @@ class StructureParser(object):
                          )
         self.update_entity()
 
+    def _init_struct(self):
+        self.STRUCT.setup_entities()
+        self.STRUCT.assign_serial_numbers()
+        self.STRUCT.renumber_models()
+
+        # keep the first model
+        if len(self.STRUCT) > 1:
+            for idx in reversed(list(range(1, len(self.STRUCT)))):
+                del self.STRUCT[idx]
+
+        self.MODEL = self.STRUCT[0]
+        self.STRUCT.remove_alternative_conformations()
+        self.STRUCT.remove_hydrogens()
+        self.STRUCT.remove_empty_chains()
+        self._update_full_sequences()
+
     def load_from_file(self, path: str):
         """
         Load model from file, default use the first model.
@@ -315,17 +321,7 @@ class StructureParser(object):
         else:
             raise ValueError("path must be files with suffixes [ .cif, .cif.gz, .pdb or .pdb.gz]")
 
-        # force to use first model when mulitple models exist
-        self.STRUCT.renumber_models()
-        if len(self.STRUCT) > 1:
-            for idx in range(1, len(self.STRUCT)):
-                del self.STRUCT[idx]
-
-        self.MODEL = self.STRUCT[0]
-        self.STRUCT.remove_alternative_conformations()
-        self.STRUCT.remove_hydrogens()
-        self.STRUCT.remove_empty_chains()
-        self._update_full_sequences()
+        self._init_struct()
         self.update_entity()
 
     def _update_full_sequences(self):
@@ -368,13 +364,24 @@ class StructureParser(object):
         return out
 
     def polymer_sequences(self, pdbx: bool = False):
+        """
+        entity sequences for polymers
+        :param pdbx:
+        :return:
+        """
         out = dict()
+        subchain_id2entity_id = self.subchain_id_to_entity_id
+        entity_dict = {ent.name: ent for ent in self.STRUCT.entities}
+
         for ch, polymer_type in self.polymer_types.items():
             polymer = self.get_chain(ch).get_polymer()
+            entity_id = subchain_id2entity_id[polymer.subchain_id()]
+            ent = entity_dict[entity_id]
+
             if pdbx:
-                s = gemmi.pdbx_one_letter_code(polymer.extract_sequence(), gemmi.sequence_kind(polymer_type))
+                s = gemmi.pdbx_one_letter_code(ent.full_sequence, gemmi.sequence_kind(polymer_type))
             else:
-                s = polymer.make_one_letter_sequence().replace("-", "")
+                s = "".join([gemmi.find_tabulated_residue(r).one_letter_code for r in ent.full_sequence]).upper()
             out[ch] = s
         return out
 
@@ -390,6 +397,14 @@ class StructureParser(object):
 
         return out
 
+    @property
+    def subchain_id_to_entity_id(self):
+        return {ch: ent.name for ent in self.STRUCT.entities for ch in ent.subchains}
+
+    @property
+    def subchain_id_to_chain_id(self):
+        return {sch.subchain_id(): chain.name for chain in self.MODEL for sch in chain.subchains()}
+
     def get_chain(self, chain_id: str):
         return self.MODEL[chain_id]
 
@@ -656,15 +671,18 @@ class StructureParser(object):
             raise ValueError("Chain %s not in found" % chain_name_2)
 
         flag = True
+        sw_name = ""
+
         while flag:
             characters = string.ascii_letters + string.digits
             sw_name = ''.join(random.choices(characters, k=4))
             if sw_name not in self.chain_ids:
                 flag = False
 
-        self.rename_chain(chain_name_1, sw_name)
-        self.rename_chain(chain_name_2, chain_name_1)
-        self.rename_chain(sw_name, chain_name_2)
+        if sw_name != "":
+            self.rename_chain(chain_name_1, sw_name)
+            self.rename_chain(chain_name_2, chain_name_1)
+            self.rename_chain(sw_name, chain_name_2)
 
     def make_one_letter_chain(self, only_uppercase: bool = True):
         uppercase_letters = list(string.ascii_uppercase)
@@ -691,7 +709,6 @@ class StructureParser(object):
 
         # not use yet
         letters_valid = [l for l in letters if l not in self.chain_ids]
-        chains2rename = [ch for ch in self.chain_ids if ch not in letters]
         mapper = {ch: letters_valid.pop() for ch in self.chain_ids if ch not in letters}
 
         for origin_name, target_name in mapper.items():

gemmi_protools/tools/align.py

@@ -10,6 +10,7 @@ from typing import Dict, Any, List, Optional
 
 import numpy as np
 from Bio.PDB import Superimposer
+
 from gemmi_protools.io.convert import gemmi2bio, bio2gemmi
 from gemmi_protools.io.reader import StructureParser
 

gemmi_protools/tools/dockq.py

@@ -11,6 +11,7 @@ from typing import List, Tuple
 
 import gemmi
 import pandas as pd
+
 from gemmi_protools.io.reader import StructureParser
 
 
@@ -22,8 +23,8 @@ def dockq_score_interface(query_model: str,
     """
     Calculate Dockq Score for an interface (partner 1 vs partner 2)
 
-    :param query_model: str or pathlib.Path
-        Path of query model, support .pdb, .pdb.gz, .cif, .cif.gz
+    :param query_model: str
+        path of query model, support .pdb, .pdb.gz, .cif, .cif.gz
     :param native_model:
     :param partner_1_mapping: a list of chain ID mapping between query and native for partner1 of the interface
         e.g. [(q chain1, n chain1), (q chain2, n chain2)]
@@ -124,4 +125,4 @@ def dockq_score_interface(query_model: str,
         else:
             score = ""
 
-        return score
+        return dict(score=score, status=msg)

gemmi_protools/tools/mesh.py

@@ -4,12 +4,15 @@
 import os
 import subprocess
 import tempfile
-from typing import Optional, List
+from collections import defaultdict
+from typing import List, Optional, Union
 
+import freesasa
 import numpy as np
 import trimesh
 from Bio.PDB import Selection
 from Bio.PDB.ResidueDepth import _get_atom_radius, _read_vertex_array
+
 from gemmi_protools import StructureParser
 from gemmi_protools import gemmi2bio
 
@@ -39,20 +42,32 @@ def get_mesh(struct_file: str, chains: Optional[List[str]] = None, MSMS: str = "
     :return:
         https://ccsb.scripps.edu/msms/downloads/
     """
+    xyz_tmp = tempfile.NamedTemporaryFile(delete=False).name
+    surface_tmp = tempfile.NamedTemporaryFile(delete=False).name
+    msms_tmp = tempfile.NamedTemporaryFile(delete=False).name
+    face_file = surface_tmp + ".face"
+    surface_file = surface_tmp + ".vert"
 
     try:
         st = StructureParser()
         st.load_from_file(struct_file)
         st.clean_structure(remove_ligand=True)
 
-        bio_st = gemmi2bio(st.STRUCT)
+        if chains is None:
+            st_p = st
+        else:
+            for ch in chains:
+                if ch not in st.chain_ids:
+                    raise ValueError("Chain %s not found (only [%s])" % (ch, " ".join(st.chain_ids)))
+            st_p = st.pick_chains(chains)
+
+        bio_st = gemmi2bio(st_p.STRUCT)
         model = bio_st[0]
 
         # Replace pdb_to_xyzr
         # Make x,y,z,radius file
         atom_list = Selection.unfold_entities(model, "A")
 
-        xyz_tmp = tempfile.NamedTemporaryFile(delete=False).name
         with open(xyz_tmp, "w") as pdb_to_xyzr:
             for atom in atom_list:
                 x, y, z = atom.coord
@@ -60,13 +75,9 @@ def get_mesh(struct_file: str, chains: Optional[List[str]] = None, MSMS: str = "
                 pdb_to_xyzr.write(f"{x:6.3f}\t{y:6.3f}\t{z:6.3f}\t{radius:1.2f}\n")
 
         # Make surface
-        surface_tmp = tempfile.NamedTemporaryFile(delete=False).name
-        msms_tmp = tempfile.NamedTemporaryFile(delete=False).name
         MSMS = MSMS + " -no_header -probe_radius 1.5 -if %s -of %s > " + msms_tmp
         make_surface = MSMS % (xyz_tmp, surface_tmp)
         subprocess.call(make_surface, shell=True)
-        face_file = surface_tmp + ".face"
-        surface_file = surface_tmp + ".vert"
         if not os.path.isfile(surface_file):
             raise RuntimeError(
                 f"Failed to generate surface file using command:\n{make_surface}"
@@ -84,12 +95,103 @@ def get_mesh(struct_file: str, chains: Optional[List[str]] = None, MSMS: str = "
         mesh.update_faces(mesh.unique_faces())
         mesh.update_faces(mesh.nondegenerate_faces())
         mesh.remove_unreferenced_vertices()
-    finally:
-        # Remove temporary files
-        for fn in [xyz_tmp, surface_tmp, msms_tmp, face_file, surface_file]:
-            try:
-                os.remove(fn)
-            except OSError:
-                pass
+
+    # Remove temporary files
+    for fn in [xyz_tmp, surface_tmp, msms_tmp, face_file, surface_file]:
+        try:
+            os.remove(fn)
+        except OSError:
+            pass
 
     return mesh
+
+
+def get_surface_residues(struct_file: str,
+                         chains: Optional[List[str]] = None,
+                         relative_sasa_cutoff: Union[int, float] = 0.15):
+    ####################
+    # check and pick
+    ####################
+    st = StructureParser()
+    st.load_from_file(struct_file)
+    st.clean_structure()
+
+    if chains is None:
+        chains = st.chain_ids
+
+    if isinstance(chains, list):
+        if len(chains) == 0:
+            raise ValueError("chains is not set")
+        else:
+            # check if chains valid
+            for ch in chains:
+                if ch not in st.chain_ids:
+                    raise ValueError("Chain %s not found" % ch)
+
+    st_p = st.pick_chains(chains)
+    # sequences = {k: s.replace("-", "").upper() for k, s in st_p.polymer_sequences().items()}
+
+    # start from 1
+    seq_num_mapper = dict()
+    for chain in st_p.MODEL:
+        for i, res in enumerate(chain):
+            key = (chain.name, str(res.seqid.num) + res.seqid.icode.strip(), res.name)
+            seq_num_mapper[key] = i + 1
+
+    # make one upper letter chain ID
+    mapper = st_p.make_one_letter_chain(only_uppercase=True)
+    mapper_r = {v: k for k, v in mapper.items()}
+
+    ####################
+    # save to pdb
+    ####################
+    with tempfile.NamedTemporaryFile(delete=True, suffix=".pdb", mode='w') as tmp_file:
+        st_p.to_pdb(tmp_file.name)
+        structure = freesasa.Structure(tmp_file.name)
+
+    result = freesasa.calc(structure)
+
+    residue_areas = result.residueAreas()
+
+    surface_residues_relative_sasa = dict()
+    surface_atoms = defaultdict(list)
+    for atom_index in range(structure.nAtoms()):
+        ch = structure.chainLabel(atom_index)
+        ch = mapper_r.get(ch, ch)
+
+        res_num = structure.residueNumber(atom_index).strip()
+        res_name = structure.residueName(atom_index)
+        atom_sasa = result.atomArea(atom_index)
+
+        res_id = (ch, res_num, res_name)
+        res_relative_total = residue_areas[ch][res_num].relativeTotal
+        if res_relative_total > relative_sasa_cutoff:
+            if res_id not in surface_residues_relative_sasa:
+                surface_residues_relative_sasa[res_id] = res_relative_total
+            if atom_sasa > 0:
+                atom_name = structure.atomName(atom_index).strip()
+                pos = structure.coord(atom_index)
+                surface_atoms[res_id].append((atom_sasa, atom_name, pos))
+
+    results = []
+    for res_id, query_atoms in surface_atoms.items():
+        seq_loc = seq_num_mapper[res_id]
+
+        query_atoms.sort(reverse=True)
+        centroid = tuple(np.array([a[2] for a in query_atoms[0:3]]).mean(axis=0).tolist())
+        results.append((res_id[0],
+                        res_id[1],
+                        res_id[2],
+                        seq_loc,
+                        centroid,
+                        surface_residues_relative_sasa[res_id]
+                        )
+                       )
+    dtype = [("chain_name", "U5"),
+             ("residue_numi", "U8"),
+             ("residue_name", "U5"),
+             ("sequential_residue_num", "i4"),
+             ("centroid", ("f4", (3,))),
+             ("relative_sasa", "f4"),
+             ]
+    return np.array(results, dtype=dtype)

gemmi_protools/tools/pdb_annot.py

@@ -12,6 +12,7 @@ from importlib.resources import files
 
 from anarci import run_anarci
 from anarci.germlines import all_germlines
+
 from gemmi_protools import StructureParser
 
 
@@ -204,10 +205,7 @@ def annotate_pdb(struct_file: str):
     st.load_from_file(struct_file)
     st.clean_structure()
 
-    subchain_id2entity_id = dict()
-    for ent in st.STRUCT.entities:
-        for ch in ent.subchains:
-            subchain_id2entity_id[ch] = ent.name
+    subchain_id2entity_id = st.subchain_id_to_entity_id
 
     # Merge sequences
     polymers = dict()

{gemmi_protools-1.0.0.dist-info → gemmi_protools-1.0.2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gemmi_protools
-Version: 1.0.0
+Version: 1.0.2
 Summary: An Enhanced tool to process PDB structures based on Gemmi
 Author: Luo Jiejian
 Author-email: Luo Jiejian <luojiejian12@mails.ucas.ac.cn>
@@ -16,6 +16,8 @@ Requires-Dist: typeguard
 Requires-Dist: numpy
 Requires-Dist: scipy
 Requires-Dist: trimesh
+Requires-Dist: joblib
+Requires-Dist: rtree
 Requires-Dist: freesasa==2.2.1
 Dynamic: author
 Dynamic: license-file
@@ -24,10 +26,8 @@ Dynamic: license-file
 
 # Install
 ```commandline
-conda create -n gemmi_protools python=3.12
-conda install -n gemmi_protools anarci hmmer -c bioconda
-conda install -n gemmi_protools dockq trimesh -c conda-forge
-conda activate gemmi_protools
+
+conda install python=3.12.9 anarci hmmer dockq trimesh rtree -c bioconda -c conda-forge
 pip install gemmi_protools
 ```
 
@@ -36,6 +36,11 @@ pip install gemmi_protools
 ## read structures
 ```commandline
 from gemmi_protools import StructureParser
+
+# load structure
 st=StructureParser()
-st.load_from_file("your.pdb")
+st.load_from_file("7mmo.cif")
+
+# get chain IDs
+print(st.chain_ids)
 ```

{gemmi_protools-1.0.0.dist-info → gemmi_protools-1.0.2.dist-info}/RECORD

@@ -6,14 +6,14 @@ gemmi_protools/data/MHC/MHC_combined.hmm.h3m,sha256=CvNMCsobQiX-wL7iB4CreNcbpnEl
 gemmi_protools/data/MHC/MHC_combined.hmm.h3p,sha256=-mK278pRedG3-KL-DtuVAQy7La9DgXg5FcP89D6X3Ck,78325
 gemmi_protools/io/__init__.py,sha256=F6e1xNT_7lZAWQgNIneH06o2qtWYrHNr_xPUPTwwx5E,29
 gemmi_protools/io/convert.py,sha256=A1i1vPgxG1LqMSUvWtegLl9LipgUQbfmKeGJ_f00UYo,3781
-gemmi_protools/io/reader.py,sha256=hUIY0YKBXDCyiWTNgfX7KsZRUxBOb-v6KYKxlWWtzEk,33238
+gemmi_protools/io/reader.py,sha256=X4onV0IVl0Q7JVH0yg2Zy-8iPIZvRPM-aaxDapawiro,33617
 gemmi_protools/tools/__init__.py,sha256=F6e1xNT_7lZAWQgNIneH06o2qtWYrHNr_xPUPTwwx5E,29
-gemmi_protools/tools/align.py,sha256=tsn8Fp-Xc9CulVyVst4uFgL6gQKVOEvoUmdgcfF8dCg,7084
-gemmi_protools/tools/dockq.py,sha256=a6i4S0O7Z2jxqQMBQilbcxNEDm35i8hzm_anvJXB6uo,4419
-gemmi_protools/tools/mesh.py,sha256=YiHPZUS-Y0gtLTbwdUXO7jS07PS5PP452LxvPTfHFJE,2986
-gemmi_protools/tools/pdb_annot.py,sha256=MHl-2BAFr__eO1ohPPLfBR17G2wPZti7Lq9UlS7AEX4,8252
-gemmi_protools-1.0.0.dist-info/licenses/LICENSE,sha256=JuQvKcgj6n11y5y6nXr9rABv3gJSswc4eTCd5WZBtSY,1062
-gemmi_protools-1.0.0.dist-info/METADATA,sha256=bRuphRjLJsZz-CmSRKau3cxi8yCPnF-E9NDavUTS1DA,1038
-gemmi_protools-1.0.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-gemmi_protools-1.0.0.dist-info/top_level.txt,sha256=P12mYJi5O5EKIn5u-RFaWxuix431CgLacSRD7rBid_U,15
-gemmi_protools-1.0.0.dist-info/RECORD,,
+gemmi_protools/tools/align.py,sha256=oKHvpeDa62zEjLkPmuyBM6avYDl3HFeJVHeRX62I2f4,7085
+gemmi_protools/tools/dockq.py,sha256=baCuO5-GZCwrlS59T5UIXogpM44OIFIfXqksqRBAb0A,4428
+gemmi_protools/tools/mesh.py,sha256=73MuJYwS_ACJI15OsrooAAhB1Ti4fM8CJSBqFOBR7LU,6537
+gemmi_protools/tools/pdb_annot.py,sha256=enATyAHq0dE8TMsKQhsSbYj-baGrI33iviJdW2R7Hv8,8157
+gemmi_protools-1.0.2.dist-info/licenses/LICENSE,sha256=JuQvKcgj6n11y5y6nXr9rABv3gJSswc4eTCd5WZBtSY,1062
+gemmi_protools-1.0.2.dist-info/METADATA,sha256=29ea2GIobnQjR6N0VQaI6MRvcH6UCG7cTpzETUUIrCE,1034
+gemmi_protools-1.0.2.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+gemmi_protools-1.0.2.dist-info/top_level.txt,sha256=P12mYJi5O5EKIn5u-RFaWxuix431CgLacSRD7rBid_U,15
+gemmi_protools-1.0.2.dist-info/RECORD,,