gemmi-protools 1.0.3__py3-none-any.whl → 1.0.4__py3-none-any.whl

gemmi_protools/io/reader.py

@@ -1,19 +1,28 @@
 import gzip
 import io
-import itertools
 import pathlib
 import random
 import string
 from collections import defaultdict
-from copy import deepcopy
 from typing import Dict, Optional, List
 
 import gemmi
 import numpy as np
 import pandas as pd
-from joblib import Parallel, delayed
 from scipy.spatial import cKDTree
 
+ATOM = [("chain_name", "U5"),
+        ("residue_num", "i4"),
+        ("residue_icode", "U3"),
+        ("residue_name", "U5"),
+        ("atom_name", "U5"),
+        ("element", "U3"),
+        ("charge", "i1"),
+        ("b_factor", "f4"),
+        ("occupancy", "f4"),
+        ("coordinate", ("f4", (3,)))
+        ]
+
 
 def is_pdb(path: str) -> bool:
     """
@@ -318,8 +327,6 @@ class StructureParser(object):
                 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()
 
@@ -774,17 +781,23 @@ class StructureParser(object):
         new_struct.INFO["source"] = src
         return new_struct
 
-    def clean_structure(self, remove_ligand=True):
+    def clean_structure(self, remove_ligand=False, remove_hydrogen=True):
         """
         Remove water by default
 
-        :param remove_ligand:
+        :param remove_ligand: bool, default False
+        :param remove_hydrogen: bool, default True
         :return:
         """
+        self.STRUCT.remove_alternative_conformations()
+
+        if remove_hydrogen:
+            self.STRUCT.remove_hydrogens()
+
         if remove_ligand:
-            self.STRUCT.remove_waters()
-        else:
             self.STRUCT.remove_ligands_and_waters()
+        else:
+            self.STRUCT.remove_waters()
 
         self.STRUCT.remove_empty_chains()
         self.update_entity()
@@ -799,11 +812,12 @@ class StructureParser(object):
                             atom.name = 'SE'
                             atom.element = gemmi.Element('Se')
 
-    def get_atoms(self, arg: str = "*"):
+    def get_atoms(self, arg: str = "*", exclude_hydrogen=False):
         """
 
         :param arg: str, "*", "/1/*//N,CA,C,O", "/1/*"
             see gemmi.Selection
+        :param exclude_hydrogen: bool, default False
         :return:
         np.ndarray
         """
@@ -814,6 +828,9 @@ class StructureParser(object):
             for chain in sel.chains(model):
                 for residue in sel.residues(chain):
                     for atom in sel.atoms(residue):
+                        if exclude_hydrogen and atom.is_hydrogen():
+                            continue
+
                         val = (chain.name,
                                residue.seqid.num,
                                residue.seqid.icode,
@@ -827,23 +844,12 @@ class StructureParser(object):
                                )
                         res.append(val)
 
-        dtype = [("chain_name", "U5"),
-                 ("residue_num", "i4"),
-                 ("residue_icode", "U3"),
-                 ("residue_name", "U5"),
-                 ("atom_name", "U5"),
-                 ("element", "U3"),
-                 ("charge", "i1"),
-                 ("b_factor", "f4"),
-                 ("occupancy", "f4"),
-                 ("coordinate", ("f4", (3,)))
-                 ]
-        return np.array(res, dtype=dtype)
-
-    def polymer_interface_residues(self,
-                                   chains_x: List[str],
-                                   chains_y: List[str],
-                                   threshold: float = 4.5):
+        return np.array(res, dtype=ATOM)
+
+    def compute_interface(self,
+                          chains_x: List[str],
+                          chains_y: List[str],
+                          threshold: float = 5.0):
         """
         Identify PPI among protein, DNA, RNA using heavy atom distances.
         :param chains_x:
@@ -860,23 +866,8 @@ class StructureParser(object):
                                  % (ch, " ".join(list(self.polymer_types.keys())))
                                  )
 
-        def ppi_atoms(struct, chains):
-            # atoms for N and O of backbone and N, O, P, S of side chains, only for PPI searching
-            protein_atoms = ['N', 'ND1', 'ND2', 'NE', 'NE1', 'NE2', 'NH1', 'NH2', 'NZ',
-                             'O', 'OD1', 'OD2', 'OE1', 'OE2', 'OG', 'OG1', 'OH',
-                             'SD', 'SG']
-            xna_atoms = ['N1', 'N2', 'N3', 'N4', 'N6', 'N7', 'N9',
-                         'O2', "O2'", "O3'", 'O4', "O4'", "O5'", 'O6',
-                         'OP1', 'OP2', 'OP3', 'P']
-            tag = "/1/%s//%s" % (",".join(chains), ",".join(protein_atoms + xna_atoms))
-            z = struct.get_atoms(tag)
-            return z
-
-        query_struct = deepcopy(self)
-        query_struct.clean_structure(remove_ligand=True)
-
-        atom_x = ppi_atoms(query_struct, chains_x)
-        atom_y = ppi_atoms(query_struct, chains_y)
+        atom_x = self.get_atoms("/1/%s" % ",".join(chains_x), exclude_hydrogen=True)
+        atom_y = self.get_atoms("/1/%s" % ",".join(chains_y), exclude_hydrogen=True)
 
         kd_tree_x = cKDTree(atom_x["coordinate"])
         kd_tree_y = cKDTree(atom_y["coordinate"])
@@ -886,38 +877,3 @@ class StructureParser(object):
         y_res = np.unique(atom_y[pairs.col][["chain_name", "residue_num", "residue_icode", "residue_name"]])
 
         return x_res, y_res
-
-    def polymer_interface_residues_all(self, ppi_threshold: float = 4.5, n_cpus: int = 4):
-        """
-        Identify PPI among protein, DNA, RNA using heavy atom distances between all chain pairs.
-
-        :param ppi_threshold:
-        :param n_cpus:
-        :return:
-        """
-        chains = list(self.polymer_types.keys())
-        ch_pairs = list(itertools.combinations(chains, r=2))
-        ch_pairs.sort()
-
-        def _run(ch_1, ch_2):
-            key = "%s/%s" % (ch_1, ch_2)
-            res_x, res_y = self.polymer_interface_residues(chains_x=[ch_1], chains_y=[ch_2], threshold=ppi_threshold)
-
-            if len(res_x) > 0:
-                vx = ["%s/%d/%s/%s" % (a, b, c.strip(), d) for a, b, c, d in res_x.tolist()]
-                vy = ["%s/%d/%s/%s" % (a, b, c.strip(), d) for a, b, c, d in res_y.tolist()]
-                return {key: [vx, vy]}
-            else:
-                return dict()
-
-        cpu2use = max(min(n_cpus, len(ch_pairs)), 1)
-
-        outputs = dict()
-        if cpu2use == 1 or len(ch_pairs) < 50:
-            for ch_1, ch_2 in ch_pairs:
-                outputs.update(_run(ch_1, ch_2))
-        else:
-            results = Parallel(n_jobs=cpu2use)(delayed(_run)(c1, c2) for c1, c2 in ch_pairs)
-            for item in results:
-                outputs.update(item)
-        return outputs

gemmi_protools/tools/pdb_annot.py

@@ -10,6 +10,7 @@ import uuid
 from collections import defaultdict
 from importlib.resources import files
 
+import numpy as np
 from anarci import run_anarci
 from anarci.germlines import all_germlines
 
@@ -37,10 +38,21 @@ def get_fv_region(in_sequence: str):
     )
 
     mapper = dict()
+    num_mapper = dict()
     for k, v in imgt_scheme.items():
         for i in range(v[0], v[1] + 1):
             mapper[i] = k
 
+            if k == "cdr1":
+                ki = 1
+            elif k == "cdr2":
+                ki = 2
+            elif k == "cdr3":
+                ki = 3
+            else:
+                ki = 0
+            num_mapper[i] = ki
+
     inputs = [("input", in_sequence)]
     _, numbered, alignment_details, _ = run_anarci(inputs, scheme="imgt", assign_germline=True)
     if numbered[0] is None:
@@ -49,6 +61,16 @@ def get_fv_region(in_sequence: str):
     outputs = []
     for cur_numbered, cur_details in zip(numbered[0], alignment_details[0]):
         aligned_sites, start, end = cur_numbered
+        # add mask
+        # 9 for not Fv region
+        # 0 for non-CDR region, 1, 2, 3 for CDR region for the current Fv
+        mask = np.full(len(in_sequence), fill_value=9, dtype=np.int8)
+        mask[start: end + 1] = 0
+        i = 0
+        for (site_num, _), site_aa in aligned_sites:
+            if site_aa != "-":
+                mask[i + start] = num_mapper[site_num]
+                i += 1
 
         # region_seq
         regions = defaultdict(list)
@@ -119,6 +141,7 @@ def get_fv_region(in_sequence: str):
                             cdr1_aa=cdr1_seq,
                             cdr2_aa=cdr2_seq,
                             cdr3_aa=cdr3_seq,
+                            mask="".join([str(i) for i in mask.tolist()])
                             )
                        )
     return outputs

{gemmi_protools-1.0.3.dist-info → gemmi_protools-1.0.4.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gemmi_protools
-Version: 1.0.3
+Version: 1.0.4
 Summary: An Enhanced tool to process PDB structures based on Gemmi
 Author: Luo Jiejian
 Author-email: Luo Jiejian <luojiejian12@mails.ucas.ac.cn>
@@ -27,7 +27,7 @@ Dynamic: license-file
 # Install
 ```commandline
 
-conda install python=3.12.9 anarci hmmer dockq trimesh rtree -c bioconda -c conda-forge
+conda install python=3.12.9 anarci hmmer dockq trimesh rtree pdbfixer -c bioconda -c conda-forge
 pip install gemmi_protools
 ```
 

{gemmi_protools-1.0.3.dist-info → gemmi_protools-1.0.4.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=-O9h5CIGPDnE1rDGMZqfApXk3_LRacNt_DaQ0zACkto,33992
+gemmi_protools/io/reader.py,sha256=drJ8WO_N8VR3KAmvEJBsHHAgZj-hzfnM3fyIV1uc0gg,32047
 gemmi_protools/tools/__init__.py,sha256=F6e1xNT_7lZAWQgNIneH06o2qtWYrHNr_xPUPTwwx5E,29
 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-1.0.3.dist-info/licenses/LICENSE,sha256=JuQvKcgj6n11y5y6nXr9rABv3gJSswc4eTCd5WZBtSY,1062
-gemmi_protools-1.0.3.dist-info/METADATA,sha256=mfEiC_hb27lnZzrW5f3EMpuEi_ri81w0xNsl1PXqVLE,1034
-gemmi_protools-1.0.3.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-gemmi_protools-1.0.3.dist-info/top_level.txt,sha256=P12mYJi5O5EKIn5u-RFaWxuix431CgLacSRD7rBid_U,15
-gemmi_protools-1.0.3.dist-info/RECORD,,
+gemmi_protools/tools/pdb_annot.py,sha256=fjEB7xmY-SmctW_lICtf2GZ63Odu9nADlyJuOen4W8o,8906
+gemmi_protools-1.0.4.dist-info/licenses/LICENSE,sha256=JuQvKcgj6n11y5y6nXr9rABv3gJSswc4eTCd5WZBtSY,1062
+gemmi_protools-1.0.4.dist-info/METADATA,sha256=HU5R2A-uwFbMcqDTcVeNPr0U3L0XrmeYVA8DihfWfiI,1043
+gemmi_protools-1.0.4.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+gemmi_protools-1.0.4.dist-info/top_level.txt,sha256=P12mYJi5O5EKIn5u-RFaWxuix431CgLacSRD7rBid_U,15
+gemmi_protools-1.0.4.dist-info/RECORD,,