1. Added get_genome_list method

2. Added get_missing_plots method
3. updated changelog
4. updated version and requirements

bbconf/_version.py

@@ -1 +1 @@
-__version__ = "0.8.1"
+__version__ = "0.9.0"

bbconf/bbagent.py

@@ -62,6 +62,17 @@ def get_stats(self) -> StatsReturn:
             genomes_number=number_of_genomes,
         )
 
+    def get_list_genomes(self) -> List[str]:
+        """
+        Get list of genomes from the database
+
+        :return: list of genomes
+        """
+        statement = select(distinct(Bed.genome_alias))
+        with Session(self.config.db_engine.engine) as session:
+            genomes = session.execute(statement).all()
+        return [result[0] for result in genomes]
+
     @cached_property
     def list_of_licenses(self) -> List[str]:
         """

bbconf/db_utils.py

@@ -339,7 +339,7 @@ class TokenizedBed(Base):
         nullable=False,
     )
     universe_id: Mapped[str] = mapped_column(
-        ForeignKey("universes.id", ondelete="CASCADE"),
+        ForeignKey("universes.id", ondelete="CASCADE", passive_deletes=True),
         primary_key=True,
         index=True,
         nullable=False,
@@ -350,7 +350,9 @@ class TokenizedBed(Base):
 
     bed: Mapped["Bed"] = relationship("Bed", back_populates="tokenized")
     universe: Mapped["Universes"] = relationship(
-        "Universes", back_populates="tokenized"
+        "Universes",
+        back_populates="tokenized",
+        passive_deletes=True,
     )
 
 

bbconf/models/bedset_models.py

@@ -1,5 +1,5 @@
-from typing import List, Union
 import datetime
+from typing import List, Union
 
 from pydantic import BaseModel, ConfigDict, model_validator
 

bbconf/modules/bedfiles.py

@@ -1,6 +1,6 @@
 import os
 from logging import getLogger
-from typing import Dict, Union
+from typing import Dict, List, Union
 
 import numpy as np
 from geniml.bbclient import BBClient
@@ -10,7 +10,7 @@
 from pydantic import BaseModel
 from qdrant_client.models import Distance, PointIdsList, VectorParams
 from sqlalchemy import and_, delete, func, select
-from sqlalchemy.orm import Session
+from sqlalchemy.orm import Session, aliased
 from tqdm import tqdm
 
 from bbconf.config_parser.bedbaseconfig import BedBaseConfig
@@ -1179,3 +1179,41 @@ def get_tokenized_link(
             bed_id=bed_id,
             universe_id=universe_id,
         )
+
+    def get_missing_plots(
+        self, plot_name: str, limit: int = 1000, offset: int = 0
+    ) -> List[str]:
+        """
+        Get list of bed files that are missing plot
+
+        :param plot_name: plot name
+        :param limit: number of results to return
+        :param offset: offset to start from
+
+        :return: list of bed file identifiers
+        """
+        if plot_name not in list(BedPlots.model_fields.keys()):
+            raise BedBaseConfError(
+                f"Plot name: {plot_name} is not valid. Valid names: {list(BedPlots.model_fields.keys())}"
+            )
+
+        with Session(self._sa_engine) as session:
+            # Alias for subquery
+            t2_alias = aliased(Files)
+
+            # Define the subquery
+            subquery = select(t2_alias).where(t2_alias.name == plot_name).subquery()
+
+            query = (
+                select(Bed.id)
+                .outerjoin(subquery, Bed.id == subquery.c.bedfile_id)
+                .where(subquery.c.bedfile_id.is_(None))
+                .limit(limit)
+                .offset(offset)
+            )
+
+            results = session.scalars(query)
+
+            results = [result for result in results]
+
+        return results

docs/changelog.md

@@ -2,6 +2,15 @@
 
 This project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html) and [Keep a Changelog](https://keepachangelog.com/en/1.0.0/) format.
 
+# [0.9.0] - 2024-11-06
+## Changed
+- Fixed bug with uploading tss dist plot\
+
+## Added
+- Added annotations to bedsets (author, source)
+- get_genome_list method to bedfiles, that lists all available genomes
+- Added method that lists all missing plots for bedfiles (get_missing_plots)
+
 # [0.8.0] - 2024-10-23
 ## Changed
 - Updated text to bed search (now using bivec)

manual_testing.py

@@ -175,9 +175,20 @@ def get_pep():
     prj
 
 
+def get_id_plots_missing():
+    from bbconf import BedBaseAgent
+
+    agent = BedBaseAgent(config="/home/bnt4me/virginia/repos/bedhost/config.yaml")
+
+    results = agent.bed.get_missing_plots("gccontent", limit=5000)
+    print(results)
+    print(agent.get_list_genomes())
+
+
 if __name__ == "__main__":
     # zarr_s3()
     # add_s3()
     # get_from_s3()
     # biocframe()
-    get_pep()
+    # get_pep()
+    get_id_plots_missing()

requirements/requirements-all.txt

@@ -1,6 +1,6 @@
 yacman >= 0.9.1
 sqlalchemy >= 2.0.0
-geniml[ml] >= 0.5.2
+geniml[ml] >= 0.5.1
 psycopg >= 3.1.15
 colorlogs
 pydantic >= 2.9.0