biofiles 0.0.5__py3-none-any.whl → 0.0.7__py3-none-any.whl

biofiles/feature.py

@@ -0,0 +1,198 @@
+from collections import deque
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Iterator, TextIO, Type
+
+from biofiles.common import Reader, Strand
+from biofiles.types.feature import Feature, Gene, ThreePrimeUTR, Exon, UTR
+
+
+@dataclass
+class FeatureDraft:
+    idx: int
+    sequence_id: str
+    source: str
+    type_: str
+    start_original: int
+    end_original: int
+    score: float | None
+    strand: Strand | None
+    phase: int | None
+    attributes: dict[str, str]
+
+    def pick_attribute(self, *keys: str) -> str | None:
+        for key in keys:
+            if (value := self.attributes.get(key, None)) is not None:
+                return value
+        return None
+
+
+@dataclass
+class FeatureDrafts:
+    drafts: deque[FeatureDraft] = field(default_factory=deque)
+    by_id: dict[str, FeatureDraft] = field(default_factory=dict)
+    # deps: dict[int, deque[int]] = field(default_factory=lambda: defaultdict(deque))
+
+    def add(self, draft: FeatureDraft) -> None:
+        self.drafts.append(draft)
+        if id_ := draft.attributes.get("ID", None):
+            self.by_id[id_] = draft
+        # if parent_id := draft.attributes.get("Parent", None):
+        #     parent = self.by_id[parent_id]
+        #     self.deps[parent.idx].append(draft.idx)
+
+    # def remove_first_n(self, n: int) -> None:
+    #     for _ in range(n):
+    #         draft = self.drafts.popleft()
+    #         if id_ := draft.attributes.get("ID", None):
+    #             del self.by_id[id_]
+    #         self.deps.pop(draft.idx, None)
+
+
+@dataclass
+class Features:
+    features: list[Feature] = field(default_factory=list)
+    by_id: dict[str, Feature] = field(default_factory=dict)
+
+    def add(self, feature: Feature):
+        self.features.append(feature)
+        if id_ := feature.id:
+            self.by_id[id_] = feature
+
+
+class FeatureReader(Reader):
+    def __init__(
+        self, input_: TextIO | Path | str, /, streaming_window: int | None = 1000
+    ):
+        super().__init__(input_)
+        self._streaming_window = streaming_window
+
+    def __iter__(self) -> Iterator[Feature]:
+        raise NotImplementedError
+
+    def _finalize_drafts(
+        self, drafts: FeatureDrafts, w: int | None
+    ) -> Iterator[Feature]:
+        # TODO streaming version!
+        #      code below is already tracking
+        # if not drafts.drafts:
+        #     return
+        # if w is not None and len(drafts.drafts) <= w:
+        #     return
+        #
+        # end_idx = drafts.drafts[-w].idx if w is not None else drafts.drafts[-1].idx
+        #
+        # i = 0
+        # while i < len(drafts.drafts) and (
+        #     not drafts.deps[drafts.drafts[i].idx]
+        #     or drafts.deps[drafts.drafts[i].idx][-1] <= end_idx
+        # ):
+        #     i += 1
+        #
+        # print(f"FINALIZING {i} DRAFTS OUT OF {len(drafts.drafts)}")
+        #
+        # result = _Features()
+        # for j in range(i):
+        #     draft = drafts.drafts[j]
+        #     feature = self._finalize_draft(draft, result)
+        #     result.add(feature)
+        # drafts.remove_first_n(i)
+        # yield from result.features
+
+        result = Features()
+        for draft in drafts.drafts:
+            feature = self._finalize_draft(draft, result)
+            result.add(feature)
+        yield from result.features
+
+    def _finalize_draft(self, draft: FeatureDraft, result: Features) -> Feature:
+        match draft.type_.lower():
+            case "gene":
+                feature = self._finalize_gene(draft, result)
+            case "exon":
+                feature = self._finalize_exon(draft, result)
+            case "three_prime_utr":
+                feature = self._finalize_utr(draft, result, ThreePrimeUTR)
+            case "utr":
+                feature = self._finalize_utr(draft, result, UTR)
+            case _:
+                feature = self._finalize_other(draft, result)
+        if feature.parent:
+            new_children = feature.parent.children + (feature,)
+            object.__setattr__(feature.parent, "children", new_children)
+        return feature
+
+    def _finalize_gene(self, draft: FeatureDraft, result: Features) -> Feature:
+        feature = self._finalize_other(draft, result)
+        name = draft.pick_attribute("gene_name", "Name")
+        biotype = draft.pick_attribute("gene_biotype", "biotype", "gene_type")
+        if name is None or biotype is None:
+            return feature
+        return Gene(**feature.__dict__, name=name, biotype=biotype, exons=())
+
+    def _finalize_exon(self, draft: FeatureDraft, result: Features) -> Feature:
+        feature = self._finalize_other(draft, result)
+
+        gene = feature.parent
+        while gene and not isinstance(gene, Gene):
+            gene = gene.parent
+
+        if gene is None:
+            return feature
+        exon = Exon(**feature.__dict__, gene=gene)
+        object.__setattr__(gene, "exons", gene.exons + (exon,))
+        return exon
+
+    def _finalize_utr(
+        self, draft: FeatureDraft, result: Features, type_: Type[UTR]
+    ) -> Feature:
+        feature = self._finalize_other(draft, result)
+
+        gene = feature.parent
+        while gene and not isinstance(gene, Gene):
+            gene = gene.parent
+
+        if gene is None:
+            return feature
+        return type_(**feature.__dict__, gene=gene)
+
+    def _finalize_other(self, draft: FeatureDraft, result: Features) -> Feature:
+        parent_id = self._extract_parent_id(draft)
+        parent = result.by_id[parent_id] if parent_id is not None else None
+
+        return Feature(
+            sequence_id=draft.sequence_id,
+            source=draft.source,
+            type_=draft.type_,
+            start_original=draft.start_original,
+            end_original=draft.end_original,
+            start_c=draft.start_original - 1,
+            end_c=draft.end_original,
+            score=draft.score,
+            strand=draft.strand,
+            phase=draft.phase,
+            attributes=draft.attributes,
+            id=self._extract_id(draft),
+            parent=parent,
+            children=(),
+        )
+
+    def _extract_id(self, draft: FeatureDraft) -> str | None:
+        if (id_ := draft.attributes.get("ID")) is not None:
+            return id_
+        if draft.type_ == "gene" and (id_ := draft.attributes.get("gene_id")):
+            return id_
+        if draft.type_ == "transcript" and (
+            id_ := draft.attributes.get("transcript_id")
+        ):
+            return id_
+        return None
+
+    def _extract_parent_id(self, draft: FeatureDraft) -> str | None:
+        if (id_ := draft.attributes.get("Parent")) is not None:
+            return id_
+        if draft.type_ == "transcript" and (id_ := draft.attributes.get("gene_id")):
+            return id_
+        if draft.type_ == "exon" and (id_ := draft.attributes.get("transcript_id")):
+            return id_
+        return None

biofiles/gff.py

@@ -1,75 +1,15 @@
 import sys
-from collections import deque
-from dataclasses import dataclass, field
 from pathlib import Path
 from typing import Iterator, cast, TextIO
 
-from biofiles.common import Strand, Reader, Writer
-from biofiles.types.feature import Feature, Gene, Exon, ThreePrimeUTR
+from biofiles.common import Strand, Writer
+from biofiles.feature import FeatureReader, FeatureDraft, FeatureDrafts
+from biofiles.types.feature import Feature, Gene, Exon
 
 __all__ = ["GFFReader", "GFF3Writer"]
 
 
-@dataclass
-class _FeatureDraft:
-    idx: int
-    sequence_id: str
-    source: str
-    type_: str
-    start_original: int
-    end_original: int
-    score: float | None
-    strand: Strand | None
-    phase: int | None
-    attributes: dict[str, str]
-
-    def pick_attribute(self, *keys: str) -> str | None:
-        for key in keys:
-            if (value := self.attributes.get(key, None)) is not None:
-                return value
-        return None
-
-
-@dataclass
-class _FeatureDrafts:
-    drafts: deque[_FeatureDraft] = field(default_factory=deque)
-    by_id: dict[str, _FeatureDraft] = field(default_factory=dict)
-    # deps: dict[int, deque[int]] = field(default_factory=lambda: defaultdict(deque))
-
-    def add(self, draft: _FeatureDraft) -> None:
-        self.drafts.append(draft)
-        if id_ := draft.attributes.get("ID", None):
-            self.by_id[id_] = draft
-        # if parent_id := draft.attributes.get("Parent", None):
-        #     parent = self.by_id[parent_id]
-        #     self.deps[parent.idx].append(draft.idx)
-
-    # def remove_first_n(self, n: int) -> None:
-    #     for _ in range(n):
-    #         draft = self.drafts.popleft()
-    #         if id_ := draft.attributes.get("ID", None):
-    #             del self.by_id[id_]
-    #         self.deps.pop(draft.idx, None)
-
-
-@dataclass
-class _Features:
-    features: list[Feature] = field(default_factory=list)
-    by_id: dict[str, Feature] = field(default_factory=dict)
-
-    def add(self, feature: Feature):
-        self.features.append(feature)
-        if id_ := feature.attributes.get("ID", None):
-            self.by_id[id_] = feature
-
-
-class GFFReader(Reader):
-    def __init__(
-        self, input_: TextIO | Path | str, /, streaming_window: int | None = 1000
-    ):
-        super().__init__(input_)
-        self._streaming_window = streaming_window
-
+class GFFReader(FeatureReader):
     def __iter__(self) -> Iterator[Feature]:
         for line in self._input:
             line = line.rstrip("\n")
@@ -84,7 +24,7 @@ class GFFReader(Reader):
             raise ValueError(f"unexpected line {line!r}, expected version")
 
     def _read_gff3(self) -> Iterator[Feature]:
-        drafts = _FeatureDrafts()
+        drafts = FeatureDrafts()
         idx = 0
         for line in self._input:
             if line.startswith("#"):
@@ -118,7 +58,7 @@ class GFFReader(Reader):
                     f"unexpected line {line!r}, parent ID not among recent feature IDs"
                 )
 
-            draft = _FeatureDraft(
+            draft = FeatureDraft(
                 idx=idx,
                 sequence_id=sequence_id,
                 source=source,
@@ -137,110 +77,6 @@ class GFFReader(Reader):
 
         yield from self._finalize_drafts(drafts, None)
 
-    def _finalize_drafts(
-        self, drafts: _FeatureDrafts, w: int | None
-    ) -> Iterator[Feature]:
-        # TODO streaming version!
-        #      code below is already tracking
-        # if not drafts.drafts:
-        #     return
-        # if w is not None and len(drafts.drafts) <= w:
-        #     return
-        #
-        # end_idx = drafts.drafts[-w].idx if w is not None else drafts.drafts[-1].idx
-        #
-        # i = 0
-        # while i < len(drafts.drafts) and (
-        #     not drafts.deps[drafts.drafts[i].idx]
-        #     or drafts.deps[drafts.drafts[i].idx][-1] <= end_idx
-        # ):
-        #     i += 1
-        #
-        # print(f"FINALIZING {i} DRAFTS OUT OF {len(drafts.drafts)}")
-        #
-        # result = _Features()
-        # for j in range(i):
-        #     draft = drafts.drafts[j]
-        #     feature = self._finalize_draft(draft, result)
-        #     result.add(feature)
-        # drafts.remove_first_n(i)
-        # yield from result.features
-
-        result = _Features()
-        for draft in drafts.drafts:
-            feature = self._finalize_draft(draft, result)
-            result.add(feature)
-        yield from result.features
-
-    def _finalize_draft(self, draft: _FeatureDraft, result: _Features) -> Feature:
-        match draft.type_.lower():
-            case "gene":
-                feature = self._finalize_gene(draft, result)
-            case "exon":
-                feature = self._finalize_exon(draft, result)
-            case "three_prime_utr":
-                feature = self._finalize_three_prime_utr(draft, result)
-            case _:
-                feature = self._finalize_other(draft, result)
-        if feature.parent:
-            new_children = feature.parent.children + (feature,)
-            object.__setattr__(feature.parent, "children", new_children)
-        return feature
-
-    def _finalize_gene(self, draft: _FeatureDraft, result: _Features) -> Feature:
-        feature = self._finalize_other(draft, result)
-        name = draft.pick_attribute("gene_name", "Name")
-        biotype = draft.pick_attribute("gene_biotype", "biotype")
-        if name is None or biotype is None:
-            return feature
-        return Gene(**feature.__dict__, name=name, biotype=biotype, exons=())
-
-    def _finalize_exon(self, draft: _FeatureDraft, result: _Features) -> Feature:
-        feature = self._finalize_other(draft, result)
-
-        gene = feature.parent
-        while gene and not isinstance(gene, Gene):
-            gene = gene.parent
-
-        if gene is None:
-            return feature
-        exon = Exon(**feature.__dict__, gene=gene)
-        object.__setattr__(gene, "exons", gene.exons + (exon,))
-        return exon
-
-    def _finalize_three_prime_utr(
-        self, draft: _FeatureDraft, result: _Features
-    ) -> Feature:
-        feature = self._finalize_other(draft, result)
-
-        gene = feature.parent
-        while gene and not isinstance(gene, Gene):
-            gene = gene.parent
-
-        if gene is None:
-            return feature
-        return ThreePrimeUTR(**feature.__dict__, gene=gene)
-
-    def _finalize_other(self, draft: _FeatureDraft, result: _Features) -> Feature:
-        parent_id = draft.attributes.get("Parent", None)
-        parent = result.by_id[parent_id] if parent_id is not None else None
-
-        return Feature(
-            sequence_id=draft.sequence_id,
-            source=draft.source,
-            type_=draft.type_,
-            start_original=draft.start_original,
-            end_original=draft.end_original,
-            start_c=draft.start_original - 1,
-            end_c=draft.end_original,
-            score=draft.score,
-            strand=draft.strand,
-            phase=draft.phase,
-            attributes=draft.attributes,
-            parent=parent,
-            children=(),
-        )
-
     def _parse_score(self, line: str, score_str: str) -> float | None:
         if score_str == ".":
             return None
@@ -286,8 +122,8 @@ class GFF3Writer(Writer):
             feature.sequence_id,
             feature.source,
             feature.type_,
-            str(feature.start_original),
-            str(feature.end_original),
+            str(feature.start_c + 1),
+            str(feature.end_c),
             str(feature.score) if feature.score is not None else ".",
             str(feature.strand) if feature.strand is not None else ".",
             str(feature.phase) if feature.phase is not None else ".",

biofiles/gtf.py

@@ -0,0 +1,38 @@
+__all__ = ["GTFReader"]
+
+import sys
+from typing import Iterator
+
+from biofiles.gff import GFFReader
+from biofiles.types.feature import Gene, Exon, Feature
+
+
+class GTFReader(GFFReader):
+    def __iter__(self) -> Iterator[Feature]:
+        yield from self._read_gff3()
+
+    def _parse_attributes(self, line: str, attributes_str: str) -> dict[str, str]:
+        return {
+            k: v.strip('"')
+            for part in attributes_str.strip(";").split(";")
+            for k, v in (part.strip().split(None, 1),)
+        }
+
+
+if __name__ == "__main__":
+    for path in sys.argv[1:]:
+        with GTFReader(path) as r:
+            total_features = 0
+            annotated_genes = 0
+            annotated_exons = 0
+            parsed_genes = 0
+            parsed_exons = 0
+            for feature in r:
+                total_features += 1
+                annotated_genes += feature.type_ == "gene"
+                annotated_exons += feature.type_ == "exon"
+                parsed_genes += isinstance(feature, Gene)
+                parsed_exons += isinstance(feature, Exon)
+        print(
+            f"{path}: {total_features} features, {parsed_genes} genes parsed out of {annotated_genes}, {parsed_exons} exons parsed out of {annotated_exons}"
+        )

biofiles/types/feature.py

@@ -14,7 +14,7 @@ class Feature:
 
     start_original: int
     end_original: int
-    # Original, 1-based inclusive values.
+    # Original values as they were present in the file (1-based inclusive for .gff and .gtf).
 
     start_c: int
     end_c: int
@@ -25,6 +25,7 @@ class Feature:
     phase: int | None
     attributes: dict[str, str]
 
+    id: str | None
     parent: "GFFFeature | None"
     children: tuple["Feature", ...]
 
@@ -46,6 +47,11 @@ class Exon(Feature):
 
 
 @dataclass(frozen=True)
-class ThreePrimeUTR(Feature):
+class UTR(Feature):
     gene: Gene
     # TODO transcript
+
+
+@dataclass(frozen=True)
+class ThreePrimeUTR(UTR):
+    pass

{biofiles-0.0.5.dist-info → biofiles-0.0.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.2
 Name: biofiles
-Version: 0.0.5
+Version: 0.0.7
 Summary: Pure-Python, zero-dependency collection of bioinformatics-related file readers and writers
 Author-email: Tigran Saluev <tigran@saluev.com>
 Maintainer-email: Tigran Saluev <tigran@saluev.com>

{biofiles-0.0.5.dist-info → biofiles-0.0.7.dist-info}/RECORD

@@ -1,14 +1,16 @@
 biofiles/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 biofiles/common.py,sha256=Yi0i85FpD2wR3vqL645LTUAE6TybGDxxZQsUmEGHqu4,1126
 biofiles/fasta.py,sha256=ctIt5I_fcZx-xQN921zpmlZS7e9_ICf-3_i6mTs5qbs,2135
-biofiles/gff.py,sha256=lSS7XxS4rpCesCc9FTG87E8p4eJYgtzN9lcz9Y5WNDk,11148
+biofiles/feature.py,sha256=wga91SYc2cMh5r5nH2m9IRM5hnVObesrxYr31kBU4WA,6874
+biofiles/gff.py,sha256=W5AjaQL_iYk4OF-H7C2pOjtpeLDEKfVg5uTOFxPDJ5I,5506
+biofiles/gtf.py,sha256=hWfjQjzwsrXLjCGr9ia6GdHNdYtlwkBrG1ldJYhRD-4,1251
 biofiles/repeatmasker.py,sha256=DqD1z1hUfCP4-qnfjF-oMF-ZpW_6XhOf_nzA8VHhQbw,3079
 biofiles/types/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-biofiles/types/feature.py,sha256=eMBrL0OQGo0MRjM7Sqmo-AKxMb0rJb1Cwc1ygp5BfPY,919
+biofiles/types/feature.py,sha256=gdywZKp7_TXiiZcXfbVs6oSHIWtju2-bqfQOmjH71Dg,1041
 biofiles/types/repeat.py,sha256=63SqzAwEGIDIGP9pxC85RUdwXbbSm0S5WNL3lSiWlmc,641
 biofiles/types/sequence.py,sha256=EOw_oKuMR0THpCYJqVE__27z7qrRqcdIPrRWTL4OFMw,152
-biofiles-0.0.5.dist-info/LICENSE,sha256=CbR8ssdFyViKj25JAlMjIt1_FbiZ1tAC5t-uwUbxqak,1070
-biofiles-0.0.5.dist-info/METADATA,sha256=cSInsFiXUGdvIK11QTD24SkhW1lWmd7DyHvd5hjE8PY,3033
-biofiles-0.0.5.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
-biofiles-0.0.5.dist-info/top_level.txt,sha256=laFaFv8hpkI4U-Pgs0yBaAJXN2_CJKl7jb-m3-tGfSc,9
-biofiles-0.0.5.dist-info/RECORD,,
+biofiles-0.0.7.dist-info/LICENSE,sha256=CbR8ssdFyViKj25JAlMjIt1_FbiZ1tAC5t-uwUbxqak,1070
+biofiles-0.0.7.dist-info/METADATA,sha256=jd6-bM7tz1laNRX_wUCnDP4FRvQwGO0sxqqWYAvowiQ,3033
+biofiles-0.0.7.dist-info/WHEEL,sha256=In9FTNxeP60KnTkGw7wk6mJPYd_dQSjEZmXdBdMCI-8,91
+biofiles-0.0.7.dist-info/top_level.txt,sha256=laFaFv8hpkI4U-Pgs0yBaAJXN2_CJKl7jb-m3-tGfSc,9
+biofiles-0.0.7.dist-info/RECORD,,