On missing gene name for an UniProtKB entry, use the AC.

translate2modelcif.py

@@ -228,6 +228,11 @@ def _abort_msg(msg, exit_code=1):
     sys.exit(exit_code)
 
 
+def _warn_msg(msg):
+    """Write a warning message to stdout."""
+    print(f"WARNING: {msg}")
+
+
 def _check_file(file_path):
     """Make sure a file exists and is actually a file."""
     if not os.path.exists(file_path):
@@ -292,9 +297,8 @@ def _parse_colabfold_config(cnfg_file):
         use_mmseqs = False
         use_msa = False
     elif cf_config["msa_mode"] == "custom":
-        print(
-            "WARNING: Custom MSA mode used. Not clear from config what to do "
-            + "here!"
+        _warn_msg(
+            "Custom MSA mode used. Not clear from config what to do here!"
         )
         seq_dbs = []
         use_mmseqs = False
@@ -336,9 +340,9 @@ def _parse_colabfold_config(cnfg_file):
     else:
         mdl_description += ", without model relaxation"
     if cf_config["use_templates"]:
-        print(
-            "WARNING: ColabFold may use PDB70 or custom templates. "
-            "Not clear from config!"
+        _warn_msg(
+            "ColabFold may use PDB70 or custom templates. Not clear "
+            + "from config!"
         )
         mdl_description += ", using templates"
     else:
@@ -557,10 +561,10 @@ def _check_sequence(up_ac, sequence):
     for i, res in enumerate(sequence):
         if res not in "ACDEFGHIKLMNPQRSTVWY":
             if res == "U":
-                print(
-                    f"WARNING: Selenocysteine found at position {i+1} of "
-                    f"entry '{up_ac}', this residue may be missing in the "
-                    "model."
+                _warn_msg(
+                    f"Selenocysteine found at position {i+1} of entry "
+                    + f"'{up_ac}', this residue may be missing in the "
+                    + "model."
                 )
                 ns_aa_pos.append(i)
                 continue
@@ -636,7 +640,11 @@ def _get_n_parse_up_entry(up_ac, up_url):
     data["up_isoform"] = None
 
     if "up_gn" not in data:
-        _abort_msg(f"No gene name found for UniProtKB entry '{up_ac}'.")
+        _warn_msg(
+            f"No gene name found for UniProtKB entry '{up_ac}', using "
+            + "UniProtKB AC instead."
+        )
+        data["up_gn"] = up_ac
     if "up_last_mod" not in data:
         _abort_msg(f"No sequence version found for UniProtKB entry '{up_ac}'.")
     if "up_crc64" not in data:
@@ -688,7 +696,7 @@ def _cmp_sequences(mdl, upkb, ns_aa_pos):
                 f"Position {pos+1} of non-canonical amino acid should be "
                 "a gap!"
             )
-        mdl = mdl[0:pos] + "U" + mdl[pos+1:]
+        mdl = mdl[0:pos] + "U" + mdl[pos + 1 :]
     return mdl == upkb