pidibble 1.2.0__tar.gz → 1.2.2__tar.gz

pidibble-1.2.2/.github/workflows/release.yaml

@@ -0,0 +1,41 @@
+name: Release & Upload to PyPI (Hatch)
+
+on:
+  push:
+    tags:
+      - "v*"
+  workflow_dispatch:
+
+jobs:
+  release:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v3
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: "3.9"
+
+      - name: Install Hatch
+        run: python -m pip install --upgrade pip hatch
+
+      - name: Clean previous builds
+        run: rm -rf dist/
+        
+      - name: Build the package with Hatch
+        run: hatch build
+
+      - name: Upload to GitHub Release
+        uses: softprops/action-gh-release@v1
+        with:
+          files: |
+            dist/*.tar.gz
+            dist/*.whl
+
+      - name: Publish to PyPI using Hatch
+        env:
+          HATCH_INDEX_USER: __token__
+          HATCH_INDEX_AUTH: ${{ secrets.PYPI_API_TOKEN }}
+        run: hatch publish

{pidibble-1.2.0 → pidibble-1.2.2}/PKG-INFO

@@ -1,6 +1,6 @@
-Metadata-Version: 2.1
+Metadata-Version: 2.4
 Name: pidibble
-Version: 1.2.0
+Version: 1.2.2
 Summary: A complete Protein Data Bank (PDB) file parser
 Project-URL: Source, https://github.com/cameronabrams/pidibble
 Project-URL: Documentation, https://pidibble.readthedocs.io/en/latest/
@@ -88,6 +88,10 @@ ATOM
 ```
 
 ## Release History
+* 1.2.2:
+   * bugfix: negative resids allowed
+* 1.2.1:
+   * bugfix: hex issues AGAIN
 * 1.2.0
    * bugfix: hex issues again
 * 1.1.9

{pidibble-1.2.0 → pidibble-1.2.2}/README.md

@@ -70,6 +70,10 @@ ATOM
 ```
 
 ## Release History
+* 1.2.2:
+   * bugfix: negative resids allowed
+* 1.2.1:
+   * bugfix: hex issues AGAIN
 * 1.2.0
    * bugfix: hex issues again
 * 1.1.9

{pidibble-1.2.0 → pidibble-1.2.2}/docs/source/conf.py

@@ -6,8 +6,8 @@ project = 'Pidibble'
 copyright = '2023-2024, Cameron F. Abrams'
 author = 'cfa22@drexel.edu'
 
-release = '1.1'
-version = '1.1.4'
+release = '1.3'
+version = '1.3.0'
 
 # -- General configuration
 

pidibble-1.2.2/docs/source/generated/pidibble.rst

@@ -0,0 +1,6 @@
+pidibble
+========
+
+.. automodule:: pidibble
+
+   

{pidibble-1.2.0 → pidibble-1.2.2}/docs/source/usage.rst

@@ -175,4 +175,6 @@ Currently, only ``ATOM``, ``HETATM``, ``SEQADV``, ``REMARK 350``, and ``REMARK 4
 >>> ', '.join(list(p.parsed.keys()))
 'ATOM, HETATM, LINK, SSBOND, SEQADV, REMARK.350.BIOMOLECULE1.TRANSFORM1, REMARK.350.BIOMOLECULE1.TRANSFORM2, REMARK.350.BIOMOLECULE1.TRANSFORM3, REMARK.465'
 
-These records are the bare minimum needed to generate (say) input coordinate and topology files for an MD simulation.  Future versions of ``pidibble`` will provide complete PDB-like parsings of ``mmCIF`` files.  This is probably not useful.
+These records are the bare minimum needed to generate (say) input coordinate and topology files for an MD simulation.  Future versions of ``pidibble`` will provide complete PDB-like parsings of ``mmCIF`` files.  This is probably not useful.
+
+Importantly:  ``pidibble`` parses mmCIF input to generate a structure that is the equivalent of the PDB format; that is, it uses ``auth`` fields instead of ``label`` fields.  

{pidibble-1.2.0 → pidibble-1.2.2}/pidibble/hex.py

@@ -4,7 +4,7 @@ def str2atomSerial(arg):
     assert type(arg)==str
     if arg=='nan':
         return_object=0
-    elif __hex_tripped__ or any([(x in arg) for x in 'abcdef']):
+    elif __hex_tripped__ or any([(x in arg) for x in 'abcdefABCDEF']):
         return_object=int(arg,16)
     elif '*' in arg:
         return_object=0

{pidibble-1.2.0 → pidibble-1.2.2}/pidibble/mmcif_parse.py

@@ -1,3 +1,4 @@
+from collections import UserDict
 from .pdbrecord import PDBRecord
 from .baserecord import BaseRecord
 import logging
@@ -28,6 +29,27 @@ def rectify(val):
         pass
     return val
 
+def resolve(key,aDict):
+    pass
+
+class MMCIFDict(UserDict):
+    def __init__(self,data,linkers={},blankers=[' ','','?']):
+        self.data=data
+        self.linkers=linkers
+        self.blankers=blankers
+    
+    def get(self,key):
+        val=self[key]
+        if val in self.blankers:
+            return ''
+        
+        key_link=self.linkers.get(val,None)
+        if key_link:
+            if key_link in self.keys():
+                val=self[key_link]
+        return val
+
+
 class MMCIF_Parser:
     def __init__(self,mmcif_formats,pdb_formats,cif_data):
         self.formats=mmcif_formats
@@ -73,6 +95,7 @@ class MMCIF_Parser:
         global_ids=mapspec.get('global_ids',{})
         spawns_on=mapspec.get('spawns_on',None)
         allcaps=mapspec.get('allcaps',[])
+        if_dot_replace_with=mapspec.get('if_dot_replace_with',{})
         cifrec=self.cif_data.getObj(mapspec['data_obj'])
         if not tables:
             for idx in range(len(cifrec)):
@@ -101,8 +124,15 @@ class MMCIF_Parser:
                                 if k in map_values:
                                     mapper=self.global_maps[map_values[k]]
                                     if type(val)==list:
-                                        val=list(set([mapper[x] for x in val]))
-                                        val.sort()
+                                        logger.debug(f'mapper {mapper}')
+                                        logger.debug(f'list before mapping {val}')
+                                        mapped_val=list(set([str(mapper[x]) for x in val]))
+                                        logger.debug(f'list after mapping {mapped_val}')
+                                        try:
+                                            mapped_val.sort()
+                                            val=mapped_val
+                                        except:
+                                            raise TypeError(f'could not sort list {mapped_val} at key {k}')
                                     else:
                                         val=mapper[val]
                         idict[k]=val

{pidibble-1.2.0 → pidibble-1.2.2}/pidibble/pdbparse.py

@@ -30,8 +30,12 @@ def safe_float(x):
 
 __version__ = importlib.metadata.version("pidibble")
 
-def str2int_sig(arg):
-    if not arg.strip().isnumeric(): return -1
+def str2int_sig(arg:str):
+    if not arg.strip().isnumeric():
+        if arg.strip()[0]=='1':
+            return int(arg)
+        else:
+            return -1
     return int(arg)
 
 class PDBParser:

{pidibble-1.2.0 → pidibble-1.2.2}/pidibble/resources/mmcif_format.yaml

@@ -9,11 +9,17 @@ ATOM:
       value: auth_asym_id
   global_ids:
     modelNum: pdbx_PDB_model_num
+  if_dot_replace_with:
+    label_seq_id: auth_seq_id
   attr_map:
     serial: id
-    name: auth_atom_id
+    name: label_atom_id
     altLoc: label_alt_id
     residue:
+      chainID: label_asym_id
+      resName: label_comp_id
+      seqNum: label_seq_id
+    residue_auth:
       chainID: auth_asym_id
       resName: auth_comp_id
       seqNum: auth_seq_id
@@ -25,10 +31,7 @@ ATOM:
     tempFactor: B_iso_or_equiv
     element: type_symbol
     charge: pdbx_formal_charge
-    label_atom_id: label_atom_id
-    label_comp_id: label_comp_id
-    label_asym_id: label_asym_id
-    label_seq_id: label_seq_id
+    auth_atom_id: auth_atom_id
 HETATM:
   data_obj: atom_site
   signal_attr: group_PDB
@@ -41,9 +44,13 @@ HETATM:
     modelNum: pdbx_PDB_model_num
   attr_map:
     serial: id
-    name: auth_atom_id
+    name: label_atom_id
     altLoc: label_alt_id
     residue:
+      chainID: label_asym_id
+      resName: label_comp_id
+      seqNum: label_seq_id
+    residue_auth:
       chainID: auth_asym_id
       resName: auth_comp_id
       seqNum: auth_seq_id
@@ -55,10 +62,42 @@ HETATM:
     tempFactor: B_iso_or_equiv
     element: type_symbol
     charge: pdbx_formal_charge
-    label_atom_id: label_atom_id
-    label_comp_id: label_comp_id
-    label_asym_id: label_asym_id
-    label_seq_id: label_seq_id
+    auth_atom_id: auth_atom_id
+# _struct_conn.id 
+# _struct_conn.conn_type_id 
+# _struct_conn.pdbx_leaving_atom_flag 
+# _struct_conn.pdbx_PDB_id 
+# _struct_conn.ptnr1_label_asym_id 
+# _struct_conn.ptnr1_label_comp_id 
+# _struct_conn.ptnr1_label_seq_id 
+# _struct_conn.ptnr1_label_atom_id 
+# _struct_conn.pdbx_ptnr1_label_alt_id 
+# _struct_conn.pdbx_ptnr1_PDB_ins_code 
+# _struct_conn.pdbx_ptnr1_standard_comp_id 
+# _struct_conn.ptnr1_symmetry 
+# _struct_conn.ptnr2_label_asym_id 
+# _struct_conn.ptnr2_label_comp_id 
+# _struct_conn.ptnr2_label_seq_id 
+# _struct_conn.ptnr2_label_atom_id 
+# _struct_conn.pdbx_ptnr2_label_alt_id 
+# _struct_conn.pdbx_ptnr2_PDB_ins_code 
+# _struct_conn.ptnr1_auth_asym_id 
+# _struct_conn.ptnr1_auth_comp_id 
+# _struct_conn.ptnr1_auth_seq_id 
+# _struct_conn.ptnr2_auth_asym_id 
+# _struct_conn.ptnr2_auth_comp_id 
+# _struct_conn.ptnr2_auth_seq_id 
+# _struct_conn.ptnr2_symmetry 
+# _struct_conn.pdbx_ptnr3_label_atom_id 
+# _struct_conn.pdbx_ptnr3_label_seq_id 
+# _struct_conn.pdbx_ptnr3_label_comp_id 
+# _struct_conn.pdbx_ptnr3_label_asym_id 
+# _struct_conn.pdbx_ptnr3_label_alt_id 
+# _struct_conn.pdbx_ptnr3_PDB_ins_code 
+# _struct_conn.details 
+# _struct_conn.pdbx_dist_value 
+# _struct_conn.pdbx_value_order 
+# _struct_conn.pdbx_role 
 LINK:
   data_obj: struct_conn
   signal_attr: conn_type_id
@@ -69,22 +108,30 @@ LINK:
     length: pdbx_dist_value
     name1: ptnr1_label_atom_id
     altLoc1: pdbx_ptnr1_label_alt_id
-    residue1:
+    residue1_auth:
       chainID: ptnr1_auth_asym_id 
       resName: ptnr1_auth_comp_id
       seqNum: ptnr1_auth_seq_id
       iCode: pdbx_ptnr1_PDB_ins_code
+    residue1:
+      chainID: ptnr1_label_asym_id 
+      resName: ptnr1_label_comp_id
+      seqNum: ptnr1_label_seq_id
     ptnr1_label_atom_id: ptnr1_label_atom_id
     ptnr1_label_comp_id: ptnr1_label_comp_id
     ptnr1_label_asym_id: ptnr1_label_asym_id
     ptnr1_label_seq_id: ptnr1_label_seq_id
     name2: ptnr2_label_atom_id
     altLoc2: pdbx_ptnr2_label_alt_id
-    residue2:
+    residue2_auth:
       chainID: ptnr2_auth_asym_id
       resName: ptnr2_auth_comp_id
       seqNum: ptnr2_auth_seq_id
       iCode: pdbx_ptnr2_PDB_ins_code
+    residue2:
+      chainID: ptnr2_label_asym_id 
+      resName: ptnr2_label_comp_id
+      seqNum: ptnr2_label_seq_id
     ptnr2_label_atom_id: ptnr2_label_atom_id
     ptnr2_label_comp_id: ptnr2_label_comp_id
     ptnr2_label_asym_id: ptnr2_label_asym_id
@@ -98,33 +145,56 @@ SSBOND:
     sym1: ptnr1_symmetry
     sym2: ptnr2_symmetry
     length: pdbx_dist_value
-    residue1:
+    residue1_auth:
       chainID: ptnr1_auth_asym_id
       resName: ptnr1_auth_comp_id
       seqNum: ptnr1_auth_seq_id
       iCode: pdbx_ptnr1_PDB_ins_code
+    residue1:
+      chainID: ptnr1_label_asym_id
+      resName: ptnr1_label_comp_id
+      seqNum: ptnr1_label_seq_id
     ptnr1_label_atom_id: ptnr1_label_atom_id
     ptnr1_label_comp_id: ptnr1_label_comp_id
     ptnr1_label_asym_id: ptnr1_label_asym_id
     ptnr1_label_seq_id: ptnr1_label_seq_id
-    residue2:
+    residue2_auth:
       chainID: ptnr2_auth_asym_id
       resName: ptnr2_auth_comp_id
       seqNum: ptnr2_auth_seq_id
       iCode: pdbx_ptnr2_PDB_ins_code
+    residue2:
+      chainID: ptnr2_label_asym_id
+      resName: ptnr2_label_comp_id
+      seqNum: ptnr2_label_seq_id
     ptnr2_label_atom_id: ptnr2_label_atom_id
     ptnr2_label_comp_id: ptnr2_label_comp_id
     ptnr2_label_asym_id: ptnr2_label_asym_id
     ptnr2_label_seq_id: ptnr2_label_seq_id
 SEQADV:
+# _struct_ref_seq_dif.align_id 
+# _struct_ref_seq_dif.pdbx_pdb_id_code 
+# _struct_ref_seq_dif.mon_id 
+# _struct_ref_seq_dif.pdbx_pdb_strand_id 
+# _struct_ref_seq_dif.seq_num 
+# _struct_ref_seq_dif.pdbx_pdb_ins_code 
+# _struct_ref_seq_dif.pdbx_seq_db_name 
+# _struct_ref_seq_dif.pdbx_seq_db_accession_code 
+# _struct_ref_seq_dif.db_mon_id 
+# _struct_ref_seq_dif.pdbx_seq_db_seq_num 
+# _struct_ref_seq_dif.details 
+# _struct_ref_seq_dif.pdbx_auth_seq_num 
+# _struct_ref_seq_dif.pdbx_ordinal 
   data_obj: struct_ref_seq_dif
   attr_map:
     idCode: pdbx_pdb_id_code
-    residue:
+    residue_auth:
       chainID: pdbx_pdb_strand_id
       resName: mon_id
       seqNum: pdbx_auth_seq_num
       iCode: pdbx_pdb_ins_code
+    residue:
+      chainID: 
     database: pdbx_seq_db_name
     dbAccession: pdbx_seq_db_accession_code
     dbRes: db_mon_id
@@ -174,15 +244,29 @@ REMARK.350:
         - 1
         - 2
         - 3
+# _pdbx_unobs_or_zero_occ_residues.id 
+# _pdbx_unobs_or_zero_occ_residues.PDB_model_num 
+# _pdbx_unobs_or_zero_occ_residues.polymer_flag 
+# _pdbx_unobs_or_zero_occ_residues.occupancy_flag 
+# _pdbx_unobs_or_zero_occ_residues.auth_asym_id 
+# _pdbx_unobs_or_zero_occ_residues.auth_comp_id 
+# _pdbx_unobs_or_zero_occ_residues.auth_seq_id 
+# _pdbx_unobs_or_zero_occ_residues.PDB_ins_code 
+# _pdbx_unobs_or_zero_occ_residues.label_asym_id 
+# _pdbx_unobs_or_zero_occ_residues.label_comp_id 
+# _pdbx_unobs_or_zero_occ_residues.label_seq_id 
 REMARK.465:
   data_obj: pdbx_unobs_or_zero_occ_residues
   tables: 
     MISSING:
       row_attr_map:
         modelNum: PDB_model_num
-        resName: auth_comp_id
-        chainID: auth_asym_id
-        seqNum: auth_seq_id
+        resName: label_comp_id
+        chainID: label_asym_id
+        seqNum: label_seq_id
+        auth_chainID: auth_asym_id
+        auth_resName: auth_comp_id
+        auth_seqNum: auth_seq_id
         iCode: PDB_ins_code
       blank_if_single_valued:
         - modelNum

{pidibble-1.2.0 → pidibble-1.2.2}/pidibble/resources/pdb_format.yaml

@@ -405,7 +405,7 @@ record_formats:
   HELIX:
     type: 3
     fields:
-      serNum: [nteger,[8,10]]
+      serNum: [Integer,[8,10]]
       helixID: [String,[12,14]]
       initRes: [Residue11,[16,26]]
       endRes: [Residue11,[28,38]]

{pidibble-1.2.0 → pidibble-1.2.2}/pyproject.toml

@@ -3,7 +3,7 @@ requires = ["hatchling"]
 build-backend = "hatchling.build"
 [project]
 name = "pidibble"
-version = "1.2.0"
+version = "1.2.2"
 authors = [
   { name="Cameron F Abrams", email="cfa22@drexel.edu" },
 ]