gffkit 0.1.0__tar.gz → 0.2__tar.gz

{gffkit-0.1.0/src/gffkit.egg-info → gffkit-0.2}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gffkit
-Version: 0.1.0
+Version: 0.2
 Summary: Region-aware GFF annotation integration toolkit
 Author: Qunjie Zhang
 License: MIT

{gffkit-0.1.0 → gffkit-0.2}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "gffkit"
-version = "0.1.0"
+version = "0.2"
 description = "Region-aware GFF annotation integration toolkit"
 readme = "README.md"
 requires-python = ">=3.8"

{gffkit-0.1.0 → gffkit-0.2}/src/gffkit/__init__.py

@@ -1,3 +1,3 @@
 """gffkit: region-aware GFF annotation integration utilities."""
 
-__version__ = "0.1.0"
+__version__ = "0.2"

{gffkit-0.1.0 → gffkit-0.2}/src/gffkit/complement_annotations.py

@@ -570,16 +570,81 @@ def cds_size_nt(root: Feature) -> int:
     return total
 
 
-def any_feature_overlap(features_a: List[Feature], features_b: List[Feature]) -> bool:
-    """判断两组特征中是否存在任意一对重叠。"""
-    for fa in features_a:
-        for fb in features_b:
-            if fa.overlaps(fb):
-                return True
-    return False
-
-
-def should_add_root(add_root: Feature, ref_roots: List[Feature], size_min: int) -> bool:
+def any_feature_overlap(features_a: List[Feature], features_b: List[Feature]) -> bool:
+    """判断两组特征中是否存在任意一对重叠。"""
+    for fa in features_a:
+        for fb in features_b:
+            if fa.overlaps(fb):
+                return True
+    return False
+
+
+def split_source_names(source: str) -> List[str]:
+    """把第二列 source 拆成有序、去重的软件名列表。"""
+    names: List[str] = []
+    for item in re.split(r"[,|+]", source):
+        name = item.strip()
+        if name and name != "." and name not in names:
+            names.append(name)
+    return names
+
+
+def collect_source_names(root: Feature) -> List[str]:
+    """收集一棵特征树中出现过的 source 名称，保持首次出现顺序。"""
+    names: List[str] = []
+    for feat in root.iter_all():
+        for name in split_source_names(feat.source):
+            if name not in names:
+                names.append(name)
+    return names
+
+
+def merge_source_names(*source_groups: Iterable[str]) -> str:
+    """合并多组 source 名称并返回可写入 GFF 第二列的字符串。"""
+    merged: List[str] = []
+    for group in source_groups:
+        for source in group:
+            for name in split_source_names(source):
+                if name not in merged:
+                    merged.append(name)
+    return ",".join(merged) if merged else "."
+
+
+def set_tree_source(root: Feature, source: str) -> None:
+    """把整棵特征树的第二列 source 统一为指定值。"""
+    for feat in root.iter_all():
+        feat.source = source
+
+
+def update_overlapping_sources(add_root: Feature, ref_roots: List[Feature]) -> List[Feature]:
+    """
+    对与 add_root 重叠的已有一级特征，合并双方软件来源并写回第二列。
+
+    返回发生重叠的参考一级特征列表，供后续补充判断复用。
+    """
+    overlapping_refs = [r for r in ref_roots if add_root.overlaps(r)]
+    if not overlapping_refs:
+        return []
+
+    add_sources = collect_source_names(add_root)
+    combined_sources: List[str] = []
+    for ref_root in overlapping_refs:
+        for source in collect_source_names(ref_root):
+            if source not in combined_sources:
+                combined_sources.append(source)
+    for source in add_sources:
+        if source not in combined_sources:
+            combined_sources.append(source)
+
+    merged_source = merge_source_names(combined_sources)
+    set_tree_source(add_root, merged_source)
+    for ref_root in overlapping_refs:
+        set_tree_source(ref_root, merged_source)
+
+    return overlapping_refs
+
+
+def should_add_root(add_root: Feature, ref_roots: List[Feature], size_min: int) -> bool:
     """
     按原脚本说明中的规则，判断 add_root 是否应该加入参考注释。
 
@@ -596,7 +661,7 @@ def should_add_root(add_root: Feature, ref_roots: List[Feature], size_min: int)
     - 只要有一个重叠关系不满足条件，就不加入整个一级特征。
       这与原说明中“只要有一个 isoform 重叠，就阻止整个 gene 加入”的精神一致。
     """
-    overlapping_refs = [r for r in ref_roots if add_root.overlaps(r)]
+    overlapping_refs = [r for r in ref_roots if add_root.overlaps(r)]
 
     # 情况 1：完全不与参考一级特征重叠
     if not overlapping_refs:
@@ -724,7 +789,7 @@ def uniquify_feature_tree(root: Feature, existing_ids: set, existing_names: set)
             existing_names.add(feat.name)
 
 
-def complement_annotations(ref_set: AnnotationSet, add_set: AnnotationSet, size_min: int) -> int:
+def complement_annotations(ref_set: AnnotationSet, add_set: AnnotationSet, size_min: int) -> int:
     """
     把 add_set 中满足条件的一级特征补充到 ref_set 中。
     返回值：实际添加的一级特征数量。
@@ -733,15 +798,16 @@ def complement_annotations(ref_set: AnnotationSet, add_set: AnnotationSet, size_
     existing_ids = ref_set.all_ids()
     existing_names = ref_set.all_names()
 
-    for add_root in sorted(
-        add_set.roots,
-        key=lambda x: (x.seqid, x.start, x.end, x.feature_type, x.id or "")
-    ):
-        if should_add_root(add_root, ref_set.roots, size_min):
-            cloned = add_root.deep_clone()
-            uniquify_feature_tree(cloned, existing_ids, existing_names)
-            ref_set.roots.append(cloned)
-            added_count += 1
+    for add_root in sorted(
+        add_set.roots,
+        key=lambda x: (x.seqid, x.start, x.end, x.feature_type, x.id or "")
+    ):
+        update_overlapping_sources(add_root, ref_set.roots)
+        if should_add_root(add_root, ref_set.roots, size_min):
+            cloned = add_root.deep_clone()
+            uniquify_feature_tree(cloned, existing_ids, existing_names)
+            ref_set.roots.append(cloned)
+            added_count += 1
 
     return added_count
 
@@ -998,4 +1064,4 @@ def main() -> int:
 
 
 if __name__ == "__main__":
-    sys.exit(main())
+    sys.exit(main())

{gffkit-0.1.0 → gffkit-0.2/src/gffkit.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: gffkit
-Version: 0.1.0
+Version: 0.2
 Summary: Region-aware GFF annotation integration toolkit
 Author: Qunjie Zhang
 License: MIT

{gffkit-0.1.0 → gffkit-0.2}/src/gffkit.egg-info/SOURCES.txt

@@ -12,4 +12,5 @@ src/gffkit.egg-info/PKG-INFO
 src/gffkit.egg-info/SOURCES.txt
 src/gffkit.egg-info/dependency_links.txt
 src/gffkit.egg-info/entry_points.txt
-src/gffkit.egg-info/top_level.txt
+src/gffkit.egg-info/top_level.txt
+tests/test_complement_sources.py

gffkit-0.2/tests/test_complement_sources.py

@@ -0,0 +1,62 @@
+from gffkit.complement_annotations import complement_annotations, parse_annotation_file
+
+
+def test_overlapping_gene_sources_are_merged(tmp_path):
+    ref_gff = tmp_path / "ref.gff3"
+    add_gff = tmp_path / "add.gff3"
+
+    ref_gff.write_text(
+        "\n".join(
+            [
+                "##gff-version 3",
+                "chr1\tEviAnn\tgene\t100\t500\t.\t+\t.\tID=gene1",
+                "chr1\tEviAnn\tmRNA\t100\t500\t.\t+\t.\tID=tx1;Parent=gene1",
+                "chr1\tEviAnn\texon\t100\t500\t.\t+\t.\tID=exon1;Parent=tx1",
+                "",
+            ]
+        ),
+        encoding="utf-8",
+    )
+    add_gff.write_text(
+        "\n".join(
+            [
+                "##gff-version 3",
+                "chr1\tANNEVO\tgene\t120\t480\t.\t+\t.\tID=gene2",
+                "chr1\tANNEVO\tmRNA\t120\t480\t.\t+\t.\tID=tx2;Parent=gene2",
+                "chr1\tANNEVO\texon\t120\t480\t.\t+\t.\tID=exon2;Parent=tx2",
+                "",
+            ]
+        ),
+        encoding="utf-8",
+    )
+
+    ref_set = parse_annotation_file(str(ref_gff))
+    add_set = parse_annotation_file(str(add_gff))
+
+    added = complement_annotations(ref_set, add_set, size_min=0)
+
+    assert added == 0
+    assert len(ref_set.roots) == 1
+    assert {feature.source for feature in ref_set.roots[0].iter_all()} == {"EviAnn,ANNEVO"}
+
+
+def test_non_overlapping_gene_keeps_single_source(tmp_path):
+    ref_gff = tmp_path / "ref.gff3"
+    add_gff = tmp_path / "add.gff3"
+
+    ref_gff.write_text(
+        "chr1\tEviAnn\tgene\t100\t200\t.\t+\t.\tID=gene1\n",
+        encoding="utf-8",
+    )
+    add_gff.write_text(
+        "chr1\tANNEVO\tgene\t400\t500\t.\t+\t.\tID=gene2\n",
+        encoding="utf-8",
+    )
+
+    ref_set = parse_annotation_file(str(ref_gff))
+    add_set = parse_annotation_file(str(add_gff))
+
+    added = complement_annotations(ref_set, add_set, size_min=0)
+
+    assert added == 1
+    assert [root.source for root in ref_set.roots] == ["EviAnn", "ANNEVO"]