mgnify-pipelines-toolkit 1.1.2__py3-none-any.whl → 1.2.0__py3-none-any.whl

mgnify_pipelines_toolkit/analysis/assembly/antismash_gff_builder.py

@@ -22,7 +22,6 @@ import pandas as pd
 
 
 def parse_args():
-
     parser = argparse.ArgumentParser()
     parser.add_argument(
         "-i", "--input", required=True, type=str, help="Input JSON from antiSMASH"
@@ -64,7 +63,6 @@ def main():
         region_name = None
 
         for feature in record["features"]:
-
             if feature["type"] == "region":
                 # Annotate region features
                 region_name = (
@@ -129,35 +127,34 @@ def main():
             cds_by_protocluster = record["modules"][
                 "antismash.detection.hmm_detection"
             ]["rule_results"]["cds_by_protocluster"]
-            if len(cds_by_protocluster) > 0:
-                for feature in cds_by_protocluster[0][1]:
-                    if "cds_name" in feature.keys():
-                        locus_tag = feature["cds_name"]
-                        as_clusters = ",".join(
-                            list(feature["definition_domains"].keys())
+
+            if not cds_by_protocluster:
+                continue
+
+            for feature in cds_by_protocluster[0][1]:
+                if locus_tag := feature.get("cds_name"):
+                    as_clusters = ",".join(list(feature["definition_domains"].keys()))
+                    if locus_tag in attributes_dict:
+                        attributes_dict[locus_tag].update(
+                            {"as_gene_clusters": as_clusters}
                         )
-                        if locus_tag in attributes_dict.keys():
-                            attributes_dict[locus_tag].update(
-                                {"as_gene_clusters": as_clusters}
-                            )
 
         if "antismash.detection.genefunctions" in record["modules"].keys():
-            for tool in record["modules"]["antismash.detection.genefunctions"]["tools"]:
-                if tool["tool"] == "smcogs":
-                    for locus_tag in tool["best_hits"]:
-                        hit_id = tool["best_hits"][locus_tag]["hit_id"].split(":")[0]
-                        hit_desc = (
-                            tool["best_hits"][locus_tag]["hit_id"]
-                            .split(":")[1]
-                            .replace(" ", "_")
-                        )
-                        score = tool["best_hits"][locus_tag]["bitscore"]
-                        e_value = tool["best_hits"][locus_tag]["evalue"]
+            gene_function_tools = record["modules"][
+                "antismash.detection.genefunctions"
+            ]["tools"]
+            if tool_data := gene_function_tools.get("smcogs"):
+
+                for locus_tag in tool_data["best_hits"]:
+                    smcog_id = tool_data["best_hits"][locus_tag]["reference_id"]
+                    smcog_description = tool_data["best_hits"][locus_tag]["description"]
+
+                    score = tool_data["best_hits"][locus_tag]["bitscore"]
+                    e_value = tool_data["best_hits"][locus_tag]["evalue"]
 
-                        smcog_note = f"smCOG:{hit_id}:{hit_desc.replace(' ', '_')}(Score:{score}%3BE-value:{e_value})"
-                        if locus_tag in attributes_dict.keys():
-                            attributes_dict[locus_tag].update({"as_notes": smcog_note})
-                        break
+                    smcog_note = f"smCOG:{smcog_id}:{smcog_description.replace(' ', '_')}(Score:{score}%3BE-value:{e_value})"
+                    if locus_tag in attributes_dict.keys():
+                        attributes_dict[locus_tag].update({"as_notes": smcog_note})
 
     attributes = [
         ";".join(f"{k}={v}" for k, v in attrib_data.items() if v)

{mgnify_pipelines_toolkit-1.1.2.dist-info → mgnify_pipelines_toolkit-1.2.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: mgnify_pipelines_toolkit
-Version: 1.1.2
+Version: 1.2.0
 Summary: Collection of scripts and tools for MGnify pipelines
 Author-email: MGnify team <metagenomics-help@ebi.ac.uk>
 License: Apache Software License 2.0
@@ -8,7 +8,7 @@ Keywords: bioinformatics,pipelines,metagenomics
 Classifier: Programming Language :: Python :: 3
 Classifier: License :: OSI Approved :: Apache Software License
 Classifier: Operating System :: OS Independent
-Requires-Python: >=3.10
+Requires-Python: >=3.11
 Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: biopython>=1.85
@@ -20,10 +20,10 @@ Requires-Dist: click<9,>=8.1.8
 Requires-Dist: pandera<0.24,>=0.23.1
 Requires-Dist: pyfastx<3,>=2.2.0
 Requires-Dist: intervaltree<4,>=3.1.0
-Provides-Extra: tests
-Requires-Dist: pytest<9,>=8.3.5; extra == "tests"
-Requires-Dist: pytest-md>=0.2.0; extra == "tests"
-Requires-Dist: pytest-workflow==2.1.0; extra == "tests"
+Provides-Extra: test
+Requires-Dist: pytest<9,>=8.3.5; extra == "test"
+Requires-Dist: pytest-md>=0.2.0; extra == "test"
+Requires-Dist: pytest-workflow==2.1.0; extra == "test"
 Provides-Extra: dev
 Requires-Dist: pre-commit>=4.2.0; extra == "dev"
 Requires-Dist: black>=25.1.0; extra == "dev"

{mgnify_pipelines_toolkit-1.1.2.dist-info → mgnify_pipelines_toolkit-1.2.0.dist-info}/RECORD

@@ -14,7 +14,7 @@ mgnify_pipelines_toolkit/analysis/amplicon/rev_comp_se_primers.py,sha256=yLpzkRJ
 mgnify_pipelines_toolkit/analysis/amplicon/standard_primer_matching.py,sha256=K6gniytuItq5WzHLi1BsaUCOdP4Zm0_ZzW2_ns7-BTI,11114
 mgnify_pipelines_toolkit/analysis/amplicon/study_summary_generator.py,sha256=epVClL10QcllL8yu7YGjx0rXNVHL2GxHi-Ek0MOjsjo,13859
 mgnify_pipelines_toolkit/analysis/assembly/add_rhea_chebi_annotation.py,sha256=NZSNY2bqs_TQyz8riDqiEFPLKcwTgzh1C7DeVHT6V8Q,4366
-mgnify_pipelines_toolkit/analysis/assembly/antismash_gff_builder.py,sha256=wXrw1B-z4hOu5oA27Vp1WYxGP2Mk6ZY4i_T5jDZgek0,6954
+mgnify_pipelines_toolkit/analysis/assembly/antismash_gff_builder.py,sha256=vZdDIcG09hulgCp0FylwHXVSGSlwl2RsDU4_xvsrUC0,6732
 mgnify_pipelines_toolkit/analysis/assembly/combined_gene_caller_merge.py,sha256=Pq-9RSt3RCxzDMQVW1VHlHF4NtpVwCWFbg2CMkvpZZc,19089
 mgnify_pipelines_toolkit/analysis/assembly/generate_gaf.py,sha256=2T4T7aXMGPac-LZUXJF3lOUzZZF50dAKkKTSaO-4idQ,3587
 mgnify_pipelines_toolkit/analysis/assembly/gff_annotation_utils.py,sha256=6gbCRlEX1eBqzFYjOt3og-961dZ--QsCJL-7l5nzg1k,33992
@@ -49,9 +49,9 @@ mgnify_pipelines_toolkit/schemas/schemas.py,sha256=pyDZvCuWbwccQF0D7c5BN1vv36wQd
 mgnify_pipelines_toolkit/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 mgnify_pipelines_toolkit/utils/fasta_to_delimited.py,sha256=lgYIR1S4crURY7C7nFtgE6QMV4u4zCNsUrVkcRnsEEo,3996
 mgnify_pipelines_toolkit/utils/get_mpt_version.py,sha256=aS9bWrC9CP7tpxoEVg6eEYt18-pmjG7fJl5Mchz4YOU,798
-mgnify_pipelines_toolkit-1.1.2.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-mgnify_pipelines_toolkit-1.1.2.dist-info/METADATA,sha256=5ByyTshLjj3V5NYnqFinX0ROLb9GmST0m_CltiUdBiY,5811
-mgnify_pipelines_toolkit-1.1.2.dist-info/WHEEL,sha256=0CuiUZ_p9E4cD6NyLD6UG80LBXYyiSYZOKDm5lp32xk,91
-mgnify_pipelines_toolkit-1.1.2.dist-info/entry_points.txt,sha256=JSjuxAr71MTeSUPPpno22wmZYgVO-gbsXfDkgWKkF7A,3533
-mgnify_pipelines_toolkit-1.1.2.dist-info/top_level.txt,sha256=xA_wC7C01V3VwuDnqwRM2QYeJJ45WtvF6LVav4tYxuE,25
-mgnify_pipelines_toolkit-1.1.2.dist-info/RECORD,,
+mgnify_pipelines_toolkit-1.2.0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+mgnify_pipelines_toolkit-1.2.0.dist-info/METADATA,sha256=uTDvoF0oYy-1ApjeygcGbjipM9ZLt1tLArKA6xDNyl4,5807
+mgnify_pipelines_toolkit-1.2.0.dist-info/WHEEL,sha256=Nw36Djuh_5VDukK0H78QzOX-_FQEo6V37m3nkm96gtU,91
+mgnify_pipelines_toolkit-1.2.0.dist-info/entry_points.txt,sha256=JSjuxAr71MTeSUPPpno22wmZYgVO-gbsXfDkgWKkF7A,3533
+mgnify_pipelines_toolkit-1.2.0.dist-info/top_level.txt,sha256=xA_wC7C01V3VwuDnqwRM2QYeJJ45WtvF6LVav4tYxuE,25
+mgnify_pipelines_toolkit-1.2.0.dist-info/RECORD,,

{mgnify_pipelines_toolkit-1.1.2.dist-info → mgnify_pipelines_toolkit-1.2.0.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (80.3.1)
+Generator: setuptools (80.7.1)
 Root-Is-Purelib: true
 Tag: py3-none-any