Add modular analysis backend config and distribution schema

bbconf/alembic/versions/845d978eac7d_added_genomic_distribution_json_plots.py

@@ -0,0 +1,60 @@
+"""Added genomic distribution json plots
+
+Revision ID: 845d978eac7d
+Revises: 8b0b706d0827
+Create Date: 2026-08-16 18:02:03.058352
+
+"""
+
+from typing import Sequence, Union
+
+import sqlalchemy as sa
+from alembic import op
+from sqlalchemy.dialects import postgresql
+
+# revision identifiers, used by Alembic.
+revision: str = "845d978eac7d"
+down_revision: Union[str, None] = "8b0b706d0827"
+branch_labels: Union[str, Sequence[str], None] = None
+depends_on: Union[str, Sequence[str], None] = None
+
+
+def upgrade() -> None:
+    """Upgrade schema."""
+
+    op.drop_column("bed", "pephub")
+    op.add_column(
+        "bed_stats",
+        sa.Column(
+            "distributions",
+            postgresql.JSONB(astext_type=sa.Text()),
+            nullable=True,
+            comment="Full distribution arrays from gtars genomicdist (JSONB)",
+        ),
+    )
+    op.add_column(
+        "bedsets",
+        sa.Column(
+            "bedset_stats",
+            postgresql.JSONB(astext_type=sa.Text()),
+            nullable=True,
+            comment="Pre-aggregated distribution statistics from gtars (JSONB)",
+        ),
+    )
+
+
+def downgrade() -> None:
+    """Downgrade schema."""
+    # ### commands auto generated by Alembic - please adjust! ###
+    op.drop_column("bedsets", "bedset_stats")
+    op.drop_column("bed_stats", "distributions")
+    op.add_column(
+        "bed",
+        sa.Column(
+            "pephub",
+            sa.BOOLEAN(),
+            autoincrement=False,
+            nullable=False,
+            comment="Whether sample was added to pephub",
+        ),
+    )

bbconf/config_parser/bedbaseconfig.py

@@ -129,19 +129,7 @@ def _read_config_file(config_path: str) -> ConfigFile:
         """
 
         _config = yacman.YAMLConfigManager.from_yaml_file(filepath=config_path).exp
-
-        config_dict = {}
-        for field_name, annotation in ConfigFile.model_fields.items():
-            try:
-                config_dict[field_name] = annotation.annotation(
-                    **_config.get(field_name)
-                )
-            except TypeError:
-                # TODO: this should be more specific
-                config_dict[field_name] = annotation.annotation()
-
-        return ConfigFile(**config_dict)
-        # return ConfigFile.from_yaml(Path(config_path))
+        return ConfigFile(**_config)
 
     @property
     def config(self) -> ConfigFile:

bbconf/config_parser/models.py

@@ -1,5 +1,6 @@
 import logging
 from pathlib import Path
+from typing import Literal
 
 from pydantic import BaseModel, ConfigDict, computed_field, field_validator
 from yacman import load_yaml
@@ -72,14 +73,14 @@ class ConfigPath(BaseModel):
 
 class AccessMethodsStruct(BaseModel):
     type: str
-    description: str = None
+    description: str | None = None
     prefix: str
 
 
 class AccessMethods(BaseModel):
-    http: AccessMethodsStruct = None
-    s3: AccessMethodsStruct = None
-    local: AccessMethodsStruct = None
+    http: AccessMethodsStruct | None = None
+    s3: AccessMethodsStruct | None = None
+    local: AccessMethodsStruct | None = None
 
 
 class ConfigS3(BaseModel):
@@ -118,13 +119,25 @@ def modify_access(self) -> bool:
         return False
 
 
+class ConfigAnalysis(BaseModel):
+    """Analysis backend configuration.
+
+    Controls which statistics engine is used for BED file analysis.
+    """
+
+    backend: Literal["r", "gtars"] = "r"
+
+    model_config = ConfigDict(extra="forbid")
+
+
 class ConfigFile(BaseModel):
     database: ConfigDB
-    qdrant: ConfigQdrant = None
+    qdrant: ConfigQdrant | None = None
     server: ConfigServer
     path: ConfigPath
-    access_methods: AccessMethods = None
-    s3: ConfigS3 = None
+    access_methods: AccessMethods | None = None
+    s3: ConfigS3 | None = None
+    analysis: ConfigAnalysis | None = ConfigAnalysis()
 
     model_config = ConfigDict(extra="allow")
 

bbconf/config_parser/utils.py

@@ -27,7 +27,7 @@ def config_analyzer(config_path: str) -> bool:
 
     _LOGGER.info(f"Analyzing the configuration file {config_path}...")
 
-    _config = yacman.YAMLConfigManager(filepath=config_path).exp
+    _config = yacman.YAMLConfigManager.from_yaml_file(filepath=config_path).exp
 
     config_dict = {}
     for field_name, annotation in ConfigFile.model_fields.items():

bbconf/db_utils.py

@@ -20,7 +20,7 @@
     select,
     text,
 )
-from sqlalchemy.dialects.postgresql import ARRAY, JSON
+from sqlalchemy.dialects.postgresql import ARRAY, JSON, JSONB
 from sqlalchemy.engine import URL, Engine, create_engine
 from sqlalchemy.event import listens_for
 from sqlalchemy.exc import IntegrityError, ProgrammingError
@@ -282,6 +282,12 @@ class BedStats(Base):
     promotercore_percentage: Mapped[Optional[float]]
     tssdist: Mapped[Optional[float]]
 
+    distributions: Mapped[Optional[dict]] = mapped_column(
+        JSONB,
+        nullable=True,
+        comment="Full distribution arrays from gtars genomicdist (JSONB)",
+    )
+
     bed: Mapped["Bed"] = relationship("Bed", back_populates="stats")
 
     __table_args__ = (
@@ -373,6 +379,11 @@ class BedSets(Base):
     bedset_standard_deviation: Mapped[Optional[dict]] = mapped_column(
         JSON, comment="Median values of the bedset"
     )
+    bedset_stats: Mapped[Optional[dict]] = mapped_column(
+        JSONB,
+        nullable=True,
+        comment="Pre-aggregated distribution statistics from gtars (JSONB)",
+    )
 
     bedfile_count: Mapped[int] = mapped_column(
         default=0, comment="Number of bedfiles in the bedset (denormalized count)"

bbconf/models/bed_models.py

@@ -74,6 +74,8 @@ class BedStatsModel(BaseModel):
     promoterprox_frequency: float | None = None
     promoterprox_percentage: float | None = None
 
+    distributions: dict | None = None
+
     model_config = ConfigDict(extra="ignore", populate_by_name=True)
 
 

bbconf/models/bedset_models.py

@@ -1,4 +1,5 @@
 import datetime
+from typing import Optional
 
 from pydantic import BaseModel, ConfigDict, model_validator
 
@@ -7,10 +8,34 @@
 
 
 class BedSetStats(BaseModel):
+    """Bedset statistics: mean/sd of scalar columns.
+
+    Populated from bedset_means and bedset_standard_deviation database columns.
+    """
+
     mean: BedStatsModel = None
     sd: BedStatsModel = None
 
 
+class BedSetDistributions(BaseModel):
+    """Collection-level aggregated distribution statistics for a bedset.
+
+    Stored in the bedset_stats JSONB database column. Populated when
+    member bed files have been processed with the gtars analysis backend.
+    """
+
+    n_files: int = 0
+    composition: Optional[dict] = None
+    scalar_summaries: Optional[dict] = None
+    tss_histogram: Optional[dict] = None
+    widths_histogram: Optional[dict] = None
+    neighbor_distances: Optional[dict] = None
+    gc_content: Optional[dict] = None
+    region_distribution: Optional[dict] = None
+    partitions: Optional[dict] = None
+    chromosome_summaries: Optional[dict] = None
+
+
 class BedSetPlots(BaseModel):
     region_commonality: FileModel = None
 
@@ -24,6 +49,7 @@ class BedSetMetadata(BaseModel):
     submission_date: datetime.datetime = None
     last_update_date: datetime.datetime = None
     statistics: BedSetStats | None = None
+    distributions: BedSetDistributions | None = None
     plots: BedSetPlots | None = None
     description: str = None
     summary: str = None

bbconf/modules/bedsets.py

@@ -420,7 +420,11 @@ def _calculate_statistics(self, bed_ids: list[str]) -> BedSetStats:
         """
 
         _LOGGER.info("Calculating bedset statistics")
-        numeric_columns = BedStatsModel.model_fields
+        numeric_columns = [
+            name
+            for name, field in BedStatsModel.model_fields.items()
+            if field.annotation in (float, float | None)
+        ]
 
         bedset_sd = {}
         bedset_mean = {}