remove smiles from BGC and AntismashBGCLoader

NPLinker · Dec 6, 2024 · 47d60db · 47d60db
1 parent 752655c
commit 47d60db
Show file tree

Hide file tree

Showing 3 changed files with 0 additions and 13 deletions.
diff --git a/src/nplinker/genomics/antismash/antismash_loader.py b/src/nplinker/genomics/antismash/antismash_loader.py
@@ -142,7 +142,6 @@ def parse_bgc_genbank(file: str | PathLike) -> BGC:
     bgc.antismash_id = antismash_id
     bgc.antismash_file = str(file)
     bgc.antismash_region = features.get("region_number")
-    bgc.smiles = features.get("smiles")
     bgc.strain = Strain(fname)
     return bgc
 
@@ -154,11 +153,4 @@ def _parse_antismash_genbank(record: SeqRecord.SeqRecord) -> dict:
             # biopython assumes region numer is a list, but it's actually an int
             features["region_number"] = feature.qualifiers.get("region_number")[0]
             features["product"] = feature.qualifiers.get("product")
-        if feature.type == "cand_cluster":
-            smiles = feature.qualifiers.get("SMILES")
-            # space is not allowed in SMILES spec
-            # biopython generates space when reading multi-line SMILES from .gbk
-            if smiles is not None:
-                smiles = tuple(i.replace(" ", "") for i in smiles)
-            features["smiles"] = smiles
     return features
diff --git a/src/nplinker/genomics/bgc.py b/src/nplinker/genomics/bgc.py
@@ -43,9 +43,6 @@ class BGC:
             see [the paper](https://doi.org/10.1186/s40793-018-0318-y).
         description: Brief description of the BGC.
             Defaults to None.
-        smiles: A tuple of SMILES formulas of the BGC's
-            products.
-            Defaults to None.
         antismash_file: The path to the antiSMASH GenBank file.
             Defaults to None.
         antismash_id: Identifier of the antiSMASH BGC, referring
@@ -80,7 +77,6 @@ def __init__(self, id: str, /, *product_prediction: str):
 
         self.mibig_bgc_class: tuple[str] | None = None
         self.description: str | None = None
-        self.smiles: tuple[str] | None = None
 
         # antismash related attributes
         self.antismash_file: str | None = None

diff --git a/tests/unit/genomics/test_antismash_loader.py b/tests/unit/genomics/test_antismash_loader.py
@@ -68,7 +68,6 @@ def test_parse_bgc_genbank():
     assert bgc.antismash_id == "NZ_AZWB01000005"
     assert bgc.antismash_file == gbk_file
     assert bgc.antismash_region == "1"
-    assert bgc.smiles == ("NC([*])C(=O)NC([*])C(=O)NC(CO)C(=O)NC(Cc1ccccc1)C(=O)NCC(=O)O",)
 
 
 def test_parse_bgc_genbank_error():