diff --git a/alembic/versions/0002_update_taxon_cols.py b/alembic/versions/0002_update_taxon_cols.py
new file mode 100644
index 0000000..67dd0fd
--- /dev/null
+++ b/alembic/versions/0002_update_taxon_cols.py
@@ -0,0 +1,52 @@
+"""Remove outdated columns from taxonomy_info table.
+
+Revision ID: 0002_update_taxon_cols
+Revises: 0001_initial_schema
+Create Date: 2026-05-11
+
+Removing columns "defined_class" (redundant with "ncbi_class"), "mito_ref" (redundant with "mitohifi_reference_species"), and
+"busco_dataset_name" (replaced by "busco_odb10_dataset_name" and "busco_odb12_dataset_name"). Also adding "hic_specimen_sample_ids"
+column to assembly table, to replace "hic_specimen_sample_id".
+"""
+
+import sqlalchemy as sa
+from sqlalchemy.dialects.postgresql import JSONB
+
+from alembic import op
+
+revision = "0002_update_taxon_cols"
+down_revision = "0001_initial_schema"
+branch_labels = None
+depends_on = None
+
+
+def upgrade() -> None:
+    op.drop_column("taxonomy_info", "defined_class")
+    op.execute(
+        sa.text(
+            """
+            UPDATE taxonomy_info
+            SET mitohifi_reference_species = COALESCE(mitohifi_reference_species, mito_ref)
+            WHERE mito_ref IS NOT NULL
+            """
+        )
+    )
+    op.drop_column("taxonomy_info", "mito_ref")
+    op.drop_column("taxonomy_info", "busco_dataset_name")
+    op.add_column("assembly", sa.Column("hic_specimen_sample_ids", JSONB, nullable=True))
+
+
+def downgrade() -> None:
+    op.add_column(
+        "taxonomy_info",
+        sa.Column("defined_class", sa.Text(), nullable=True),
+    )
+    op.add_column(
+        "taxonomy_info",
+        sa.Column("mito_ref", sa.Text(), nullable=True),
+    )
+    op.add_column(
+        "taxonomy_info",
+        sa.Column("busco_dataset_name", sa.Text(), nullable=True),
+    )
+    op.drop_column("assembly", "hic_specimen_sample_ids")
diff --git a/app/api/v1/endpoints/assemblies.py b/app/api/v1/endpoints/assemblies.py
index 76802f7..2ddc99c 100644
--- a/app/api/v1/endpoints/assemblies.py
+++ b/app/api/v1/endpoints/assemblies.py
@@ -323,7 +323,13 @@ def get_assembly_manifest(
             detail="No manifest stored for this assembly. Re-submit an intent to generate one.",
         )
 
-    return JSONResponse(content=assembly.manifest_json)
+    return JSONResponse(
+        content={
+            "assembly_id": str(assembly.id),
+            "version": assembly.version,
+            "manifest": assembly.manifest_json,
+        }
+    )
 
 
 @router.post("/intent/{taxon_id}")
@@ -360,12 +366,15 @@ def create_assembly_intent(
         db, intent_in.long_read_specimen_sample_id, org_taxon_id, "long_read_specimen_sample_id"
     )
 
-    # 3. Validate hic_specimen_sample_id if provided
-    hic_sample = None
-    if intent_in.hic_specimen_sample_id:
-        hic_sample = _validate_specimen_sample(
-            db, intent_in.hic_specimen_sample_id, org_taxon_id, "hic_specimen_sample_id"
-        )
+    # 3. Validate hic_specimen_sample_ids if provided
+    hic_samples: List[Sample] = []
+    if intent_in.hic_specimen_sample_ids:
+        for idx, hic_sid in enumerate(intent_in.hic_specimen_sample_ids):
+            hic_samples.append(
+                _validate_specimen_sample(
+                    db, hic_sid, org_taxon_id, f"hic_specimen_sample_ids[{idx}]"
+                )
+            )
 
     # 4. Fetch long-read experiments (PacBio or ONT only) and their reads
     long_read_lineage_sample_ids = _get_lineage_sample_ids_for_specimen(db, long_read_sample)
@@ -391,16 +400,16 @@ def create_assembly_intent(
     long_read_exp_ids = [e.id for e in long_read_experiments]
     long_reads = db.query(Read).filter(Read.experiment_id.in_(long_read_exp_ids)).all()
 
-    # 5. Fetch Hi-C experiments and reads if hic_sample is provided
+    # 5. Fetch Hi-C experiments and reads for each hic_sample
     hic_experiments: List[Experiment] = []
     hic_reads: List[Read] = []
     hic_sample_id_map: Dict[str, str] = {}
-    if hic_sample:
+    for hic_sample in hic_samples:
         hic_lineage_sample_ids = _get_lineage_sample_ids_for_specimen(db, hic_sample)
-        hic_sample_id_map = {
-            str(sample_id): str(hic_sample.id) for sample_id in hic_lineage_sample_ids
-        }
-        hic_experiments = (
+        hic_sample_id_map.update(
+            {str(sample_id): str(hic_sample.id) for sample_id in hic_lineage_sample_ids}
+        )
+        hic_exps = (
             db.query(Experiment)
             .filter(
                 Experiment.sample_id.in_(hic_lineage_sample_ids),
@@ -408,18 +417,17 @@ def create_assembly_intent(
             )
             .all()
         )
-        hic_experiments = [
-            e for e in hic_experiments if (e.library_strategy or "").upper() == "HI-C"
-        ]
-        if not hic_experiments:
+        hic_exps = [e for e in hic_exps if (e.library_strategy or "").upper() == "HI-C"]
+        if not hic_exps:
             raise AppError(
                 status_code=422,
                 code="no_hic_experiments",
-                message="No Hi-C experiments (ILLUMINA + Hi-C) found for hic_specimen_sample_id",
+                message="No Hi-C experiments (ILLUMINA + Hi-C) found for hic_specimen_sample_ids",
                 details={"hic_specimen_sample_id": str(hic_sample.id)},
             )
-        hic_exp_ids = [e.id for e in hic_experiments]
-        hic_reads = db.query(Read).filter(Read.experiment_id.in_(hic_exp_ids)).all()
+        hic_exp_ids = [e.id for e in hic_exps]
+        hic_experiments.extend(hic_exps)
+        hic_reads.extend(db.query(Read).filter(Read.experiment_id.in_(hic_exp_ids)).all())
 
     # 6. Determine data_types from the relevant experiments
     # TODO review and remove this step now that we pass in the specimen_id
@@ -439,7 +447,7 @@ def create_assembly_intent(
         db,
         taxon_id=org_taxon_id,
         long_read_specimen_sample_id=long_read_sample.id,
-        hic_specimen_sample_id=hic_sample.id if hic_sample else None,
+        hic_specimen_sample_ids=[s.id for s in hic_samples] if hic_samples else None,
         # TODO review if we need data_types when creating experiment
         data_types=data_types,
         tol_id=intent_in.tol_id,
@@ -450,7 +458,7 @@ def create_assembly_intent(
     try:
         # 8. Build sample metadata and generate JSON manifest
         all_reads = long_reads + hic_reads
-        sample_metadata_by_id = _build_specimen_metadata(long_read_sample, hic_sample)
+        sample_metadata_by_id = _build_specimen_metadata(long_read_sample, *hic_samples)
         sequencing_sample_to_specimen_sample_id = {
             **long_read_sample_id_map,
             **hic_sample_id_map,
@@ -462,16 +470,19 @@ def create_assembly_intent(
             experiments=all_experiments,
             # TODO tolid from sample (reported by broker) not input from caller
             tol_id=assembly.tol_id,
+            assembly_id=str(assembly.id),
             version=assembly.version,
             long_read_sample_id=long_read_sample.id,
-            hic_sample_id=hic_sample.id if hic_sample else None,
+            hic_sample_ids=[s.id for s in hic_samples] if hic_samples else None,
             sample_metadata_by_id=sample_metadata_by_id,
             sequencing_sample_to_specimen_sample_id=sequencing_sample_to_specimen_sample_id,
         )
 
         # 9. Validate that the manifest contains eligible reads then persist it
+        # TODO decide whether to keep this validation.
+        """
         long_read_keys = {"PACBIO_SMRT", "OXFORD_NANOPORE"}
-        if not any(k in manifest_data["reads"] for k in long_read_keys):
+        if not any(k in manifest_data["read_files"] for k in long_read_keys):
             raise AppError(
                 status_code=422,
                 code="no_eligible_long_reads",
@@ -482,14 +493,14 @@ def create_assembly_intent(
                 details={"long_read_specimen_sample_id": str(long_read_sample.id)},
             )
 
-        if hic_sample and "Hi-C" not in manifest_data["reads"]:
+        if hic_samples and "Hi-C" not in manifest_data["read_files"]:
             raise AppError(
                 status_code=422,
                 code="no_eligible_hic_reads",
-                message="No eligible Hi-C reads found for hic_specimen_sample_id",
-                details={"hic_specimen_sample_id": str(hic_sample.id)},
+                message="No eligible Hi-C reads found for hic_specimen_sample_ids",
+                details={"hic_specimen_sample_ids": [str(s.id) for s in hic_samples]},
             )
-
+        """
         assembly.manifest_json = manifest_data
         db.add(assembly)
         db.commit()
@@ -503,7 +514,6 @@ def create_assembly_intent(
         content={
             "assembly_id": str(assembly.id),
             "version": assembly.version,
-            "status": assembly.status,
             "manifest": manifest_data,
         }
     )
@@ -556,9 +566,7 @@ def cancel_assembly_intent(
         "long_read_specimen_sample_id": str(assembly.long_read_specimen_sample_id)
         if assembly.long_read_specimen_sample_id
         else None,
-        "hic_specimen_sample_id": str(assembly.hic_specimen_sample_id)
-        if assembly.hic_specimen_sample_id
-        else None,
+        "hic_specimen_sample_ids": assembly.hic_specimen_sample_ids or [],
         "version": assembly.version,
         "status": assembly.status,
     }
@@ -635,7 +643,13 @@ def get_manifest_by_assembly_id(
             detail="No manifest stored for this assembly. Re-submit an intent to generate one.",
         )
 
-    return JSONResponse(content=assembly.manifest_json)
+    return JSONResponse(
+        content={
+            "assembly_id": str(assembly.id),
+            "version": assembly.version,
+            "manifest": assembly.manifest_json,
+        }
+    )
 
 
 @router.get("/{assembly_id}", response_model=AssemblySchema)
diff --git a/app/api/v1/endpoints/taxonomy_info.py b/app/api/v1/endpoints/taxonomy_info.py
index 8639694..b64e968 100644
--- a/app/api/v1/endpoints/taxonomy_info.py
+++ b/app/api/v1/endpoints/taxonomy_info.py
@@ -8,7 +8,11 @@
 from app.core.policy import policy
 from app.models.taxonomy_info import TaxonomyInfo
 from app.models.user import User
-from app.schemas.bulk_import import BulkImportResponse, BulkTaxonomyInfoImport
+from app.schemas.bulk_import import (
+    BulkImportResponse,
+    BulkNcbiRefreshRequest,
+    BulkTaxonomyInfoImport,
+)
 from app.schemas.taxonomy_info import (
     TaxonomyInfo as TaxonomyInfoSchema,
 )
@@ -50,6 +54,40 @@ def bulk_import_taxonomy_info(
     return taxonomy_info_service.bulk_import(db, data=data.root)
 
 
+@router.post("/bulk-upsert", response_model=BulkImportResponse)
+@policy("taxonomy_info:bulk_upsert")
+def bulk_upsert_taxonomy_info(
+    *,
+    db: Session = Depends(get_db),
+    data: BulkTaxonomyInfoImport,
+    current_user: User = Depends(get_current_active_user),
+) -> Any:
+    """
+    Bulk upsert taxonomy info from a dictionary keyed by taxon_id.
+
+    Inserts new rows and updates existing ones. Re-fetches NCBI data for all rows.
+    The taxon_id key must reference an existing organism.
+    """
+    return taxonomy_info_service.bulk_upsert(db, data=data.root)
+
+
+@router.post("/bulk-ncbi-refresh", response_model=BulkImportResponse)
+@policy("taxonomy_info:bulk_ncbi_refresh")
+def bulk_ncbi_refresh_taxonomy_info(
+    *,
+    db: Session = Depends(get_db),
+    data: BulkNcbiRefreshRequest,
+    current_user: User = Depends(get_current_active_user),
+) -> Any:
+    """
+    Re-sync NCBI taxonomy fields for existing taxonomy_info rows.
+
+    Only updates ncbi_* columns — upstream fields are left unchanged.
+    Rows that do not yet have a taxonomy_info record are skipped.
+    """
+    return taxonomy_info_service.bulk_ncbi_refresh(db, taxon_ids=data.taxon_ids)
+
+
 @router.get("/{taxon_id}", response_model=TaxonomyInfoSchema)
 def get_taxonomy_info(
     *,
diff --git a/app/core/policy.py b/app/core/policy.py
index 238fbb7..22524ad 100644
--- a/app/core/policy.py
+++ b/app/core/policy.py
@@ -66,6 +66,8 @@
     "taxonomy_info:update": ["curator", "admin"],
     "taxonomy_info:delete": ["admin", "superuser"],
     "taxonomy_info:bulk_import": ["curator", "admin"],
+    "taxonomy_info:bulk_upsert": ["curator", "admin"],
+    "taxonomy_info:bulk_ncbi_refresh": ["curator", "admin"],
 }
 
 
diff --git a/app/models/assembly.py b/app/models/assembly.py
index 4198d2a..7ff016a 100644
--- a/app/models/assembly.py
+++ b/app/models/assembly.py
@@ -74,6 +74,7 @@ class Assembly(Base):
         UUID(as_uuid=True), ForeignKey("sample.id"), nullable=True
     )
     hic_specimen_sample_id = Column(UUID(as_uuid=True), ForeignKey("sample.id"), nullable=True)
+    hic_specimen_sample_ids = Column(JSONB, nullable=True)
     manifest_json = Column(JSONB, nullable=True)
 
     # Relationships
diff --git a/app/models/taxonomy_info.py b/app/models/taxonomy_info.py
index 0263908..6e024c2 100644
--- a/app/models/taxonomy_info.py
+++ b/app/models/taxonomy_info.py
@@ -23,8 +23,6 @@ class TaxonomyInfo(Base):
     ncbi_tax_string = Column(Text, nullable=True)
     ncbi_full_lineage = Column(Text, nullable=True)
     ncbi_last_synced_at = Column(DateTime(timezone=True), nullable=True)
-    mito_ref = Column(Text, nullable=True)
-    busco_dataset_name = Column(Text, nullable=True)
     busco_odb10_dataset_name = Column(Text, nullable=True)
     busco_odb12_dataset_name = Column(Text, nullable=True)
     find_plastid = Column(Boolean, nullable=True)
@@ -32,7 +30,6 @@ class TaxonomyInfo(Base):
     mitochondrial_genetic_code_id = Column(Integer, nullable=True)
     mitohifi_reference_species = Column(Text, nullable=True)
     oatk_hmm_name = Column(Text, nullable=True)
-    defined_class = Column(Text, nullable=True)
     augustus_dataset_name = Column(Text, nullable=True)
     genetic_code_id = Column(Integer, nullable=True)
 
diff --git a/app/schemas/assembly.py b/app/schemas/assembly.py
index 4eb5120..7e4653f 100644
--- a/app/schemas/assembly.py
+++ b/app/schemas/assembly.py
@@ -86,7 +86,7 @@ class AssemblyIntent(BaseModel):
 
     tol_id: Optional[str] = None
     long_read_specimen_sample_id: UUID
-    hic_specimen_sample_id: Optional[UUID] = None
+    hic_specimen_sample_ids: Optional[List[UUID]] = None
 
 
 class AssemblyIntentResponse(BaseModel):
diff --git a/app/schemas/bulk_import.py b/app/schemas/bulk_import.py
index 7e99051..fb7a3ff 100644
--- a/app/schemas/bulk_import.py
+++ b/app/schemas/bulk_import.py
@@ -42,10 +42,17 @@ class BulkTaxonomyInfoImport(RootModel[Dict[int, TaxonomyInfoUpdate]]):
     """Bulk taxonomy info import keyed by taxon_id."""
 
 
+class BulkNcbiRefreshRequest(BaseModel):
+    """Request body for bulk NCBI taxonomy refresh."""
+
+    taxon_ids: List[int]
+
+
 class BulkImportResponse(BaseModel):
     """Schema for bulk import response."""
 
     created_count: int
+    updated_count: Optional[int] = 0
     skipped_count: Optional[int] = 0
     message: str
     errors: Optional[List[str]] = None  # List of all errors with context
diff --git a/app/schemas/taxonomy_info.py b/app/schemas/taxonomy_info.py
index 43468eb..f5c8602 100644
--- a/app/schemas/taxonomy_info.py
+++ b/app/schemas/taxonomy_info.py
@@ -17,8 +17,6 @@ class TaxonomyInfoBase(BaseModel):
     ncbi_tax_string: Optional[str] = None
     ncbi_full_lineage: Optional[str] = None
     ncbi_last_synced_at: Optional[datetime] = None
-    mito_ref: Optional[str] = None
-    busco_dataset_name: Optional[str] = None
     busco_odb10_dataset_name: Optional[str] = None
     busco_odb12_dataset_name: Optional[str] = None
     find_plastid: Optional[bool] = None
@@ -26,21 +24,17 @@ class TaxonomyInfoBase(BaseModel):
     mitochondrial_genetic_code_id: Optional[int] = None
     mitohifi_reference_species: Optional[str] = None
     oatk_hmm_name: Optional[str] = None
-    defined_class: Optional[str] = None
     augustus_dataset_name: Optional[str] = None
     genetic_code_id: Optional[int] = None
 
 
 class TaxonomyInfoWriteBase(BaseModel):
-    busco_dataset_name: Optional[str] = None
     busco_odb10_dataset_name: Optional[str] = None
     busco_odb12_dataset_name: Optional[str] = None
     find_plastid: Optional[bool] = None
     hic_motif: Optional[str] = None
     mitochondrial_genetic_code_id: Optional[int] = None
-    mitohifi_reference_species: Optional[str] = None
     oatk_hmm_name: Optional[str] = None
-    defined_class: Optional[str] = None
     augustus_dataset_name: Optional[str] = None
     genetic_code_id: Optional[int] = None
 
diff --git a/app/services/assembly_helper.py b/app/services/assembly_helper.py
index 9bc5698..7794444 100644
--- a/app/services/assembly_helper.py
+++ b/app/services/assembly_helper.py
@@ -128,9 +128,10 @@ def generate_assembly_manifest_json(
     reads: List[Read],
     experiments: List[Experiment],
     tol_id: str | None,
+    assembly_id: str,
     version: int,
     long_read_sample_id: UUID,
-    hic_sample_id: Optional[UUID] = None,
+    hic_sample_ids: Optional[List[UUID]] = None,
     sample_metadata_by_id: Dict[str, Dict[str, Any]] | None = None,
     sequencing_sample_to_specimen_sample_id: Dict[str, str] | None = None,
 ) -> Dict[str, Any]:
@@ -138,7 +139,7 @@ def generate_assembly_manifest_json(
 
     Reads are routed to sections based on which specimen sample they belong to:
     - Reads from long_read_sample_id experiments → PACBIO_SMRT or OXFORD_NANOPORE section
-    - Reads from hic_sample_id experiments → Hi-C section (omitted when hic_sample_id is None)
+    - Reads from hic_sample_ids experiments → Hi-C section (omitted when hic_sample_ids is None)
 
     PacBio filtering: only files ending in .ccs.bam or hifi_reads.bam are included.
     ONT: all reads are included.
@@ -147,12 +148,12 @@ def generate_assembly_manifest_json(
     Args:
         organism: Organism object
         taxonomy_info: TaxonomyInfo object related to the organism when available
-        reads: Combined list of Read objects from both specimen samples
-        experiments: Combined list of Experiment objects from both specimen samples
+        reads: Combined list of Read objects from all specimen samples
+        experiments: Combined list of Experiment objects from all specimen samples
         tol_id: ToL ID for the assembly (optional)
         version: Assembly version number
         long_read_sample_id: sample.id of the long-read specimen sample
-        hic_sample_id: sample.id of the Hi-C specimen sample (optional)
+        hic_sample_ids: sample.ids of the Hi-C specimen samples (optional)
         sample_metadata_by_id: Sample metadata keyed by sample.id as string (optional)
 
     Returns:
@@ -166,7 +167,7 @@ def generate_assembly_manifest_json(
     logger.info("Total experiments: %d, Total reads: %d", len(experiments), len(reads))
 
     long_read_sample_str = str(long_read_sample_id)
-    hic_sample_str = str(hic_sample_id) if hic_sample_id else None
+    hic_sample_id_strs = {str(sid) for sid in hic_sample_ids} if hic_sample_ids else set()
 
     # Build experiment info map: experiment.id → metadata
     exp_info_map: Dict[Any, Dict[str, Any]] = {}
@@ -208,30 +209,32 @@ def generate_assembly_manifest_json(
 
         # Route by specimen sample, then by platform
         is_long_read_sample = resolved_sample_id == long_read_sample_str
-        is_hic_sample = hic_sample_str is not None and resolved_sample_id == hic_sample_str
+        is_hic_sample = bool(hic_sample_id_strs) and resolved_sample_id in hic_sample_id_strs
 
         if is_long_read_sample:
             if platform == "PACBIO_SMRT" and library_strategy in ("WGS", "WGA") and read.file_name:
-                if read.file_name.endswith(".ccs.bam") or read.file_name.endswith("hifi_reads.bam"):
-                    logger.info("Adding PacBio read: %s", read.file_name)
-                    if bpa_package_id not in pacbio_by_package:
-                        entry: Dict[str, Any] = {
-                            "sample_id": resolved_sample_id,
-                            "bpa_sample_id": sample_meta.get("bpa_sample_id"),
-                            "specimen_id": sample_meta.get("specimen_id"),
-                            "resources": [],
-                        }
-                        if exp_info["bioplatforms_base_url"]:
-                            entry["bioplatforms_base_url"] = exp_info["bioplatforms_base_url"]
-                        pacbio_by_package[bpa_package_id] = entry
-                    pacbio_by_package[bpa_package_id]["resources"].append(
-                        {"md5sum": read.file_checksum, "url": read.bioplatforms_url}
-                    )
+                # if read.file_name.endswith(".ccs.bam") or read.file_name.endswith("hifi_reads.bam"):
+                logger.info("Adding PacBio read: %s", read.file_name)
+                if bpa_package_id not in pacbio_by_package:
+                    entry: Dict[str, Any] = {
+                        "sample_id": resolved_sample_id,
+                        "bpa_sample_id": sample_meta.get("bpa_sample_id"),
+                        "specimen_id": sample_meta.get("specimen_id"),
+                        "single_end": [],
+                    }
+                    if exp_info["bioplatforms_base_url"]:
+                        entry["bioplatforms_base_url"] = exp_info["bioplatforms_base_url"]
+                    pacbio_by_package[bpa_package_id] = entry
+                pacbio_by_package[bpa_package_id]["single_end"].append(
+                    {"md5sum": read.file_checksum, "url": read.bioplatforms_url}
+                )
+                """
                 else:
                     logger.debug(
                         "Skipping PacBio read %s — not .ccs.bam or hifi_reads.bam",
                         read.file_name,
                     )
+                """
             elif platform == "OXFORD_NANOPORE" and library_strategy in ("WGS", "WGA"):
                 logger.info("Adding ONT read: %s", read.file_name)
                 if bpa_package_id not in ont_by_package:
@@ -239,12 +242,12 @@ def generate_assembly_manifest_json(
                         "sample_id": resolved_sample_id,
                         "bpa_sample_id": sample_meta.get("bpa_sample_id"),
                         "specimen_id": sample_meta.get("specimen_id"),
-                        "resources": [],
+                        "single_end": [],
                     }
                     if exp_info["bioplatforms_base_url"]:
                         entry["bioplatforms_base_url"] = exp_info["bioplatforms_base_url"]
                     ont_by_package[bpa_package_id] = entry
-                ont_by_package[bpa_package_id]["resources"].append(
+                ont_by_package[bpa_package_id]["single_end"].append(
                     {"md5sum": read.file_checksum, "url": read.bioplatforms_url}
                 )
 
@@ -265,11 +268,12 @@ def generate_assembly_manifest_json(
                         "sample_id": resolved_sample_id,
                         "bpa_sample_id": sample_meta.get("bpa_sample_id"),
                         "specimen_id": sample_meta.get("specimen_id"),
-                        "resources": {"r1": [], "r2": []},
+                        "r1": [],
+                        "r2": [],
                     }
                 rkey = _normalize_read_number(read.read_number)
                 if rkey in ("r1", "r2"):
-                    hic_by_package[bpa_package_id]["resources"][rkey].append(
+                    hic_by_package[bpa_package_id][rkey].append(
                         {
                             "url": read.bioplatforms_url,
                             "md5sum": read.file_checksum,
@@ -314,7 +318,8 @@ def generate_assembly_manifest_json(
     manifest = {
         "scientific_name": organism.scientific_name,
         "taxon_id": organism.taxon_id,
-        "tolid": tol_id,
+        "dataset_id": tol_id,
+        "assembly_id": assembly_id,
         "version": version,
         "busco_odb10_dataset_name": getattr(taxonomy_info, "busco_odb10_dataset_name", None),
         "busco_odb12_dataset_name": getattr(taxonomy_info, "busco_odb12_dataset_name", None),
@@ -327,7 +332,7 @@ def generate_assembly_manifest_json(
         "oatk_hmm_name": getattr(taxonomy_info, "oatk_hmm_name", None),
         "augustus_dataset_name": getattr(taxonomy_info, "augustus_dataset_name", None),
         "genetic_code_id": getattr(taxonomy_info, "genetic_code_id", None),
-        "defined_class": getattr(taxonomy_info, "defined_class", None),
-        "reads": reads_section,
+        "ncbi_class": getattr(taxonomy_info, "ncbi_class", None),
+        "read_files": reads_section,
     }
     return manifest
diff --git a/app/services/assembly_service.py b/app/services/assembly_service.py
index a72ee17..432d8d3 100644
--- a/app/services/assembly_service.py
+++ b/app/services/assembly_service.py
@@ -212,7 +212,7 @@ def create_from_intent(
         *,
         taxon_id: int,
         long_read_specimen_sample_id: UUID,
-        hic_specimen_sample_id: Optional[UUID],
+        hic_specimen_sample_ids: Optional[List[UUID]],
         data_types: str,
         tol_id: Optional[str],
         project_id: Optional[UUID],
@@ -221,7 +221,7 @@ def create_from_intent(
         """Create an Assembly at manifest-request time with status='requested'.
 
         Versioning is scoped by (taxon_id, long_read_specimen_sample_id) only —
-        data_types and hic_specimen_sample_id are excluded from the version key.
+        data_types and hic_specimen_sample_ids are excluded from the version key.
         sample_id is set to long_read_specimen_sample_id for backward compatibility.
         """
         version = self.get_next_version_for_intent(
@@ -229,11 +229,16 @@ def create_from_intent(
             taxon_id=taxon_id,
             long_read_specimen_sample_id=long_read_specimen_sample_id,
         )
+        # Keep singular FK column pointing to the first HiC sample for backwards compatibility
+        hic_specimen_sample_id = hic_specimen_sample_ids[0] if hic_specimen_sample_ids else None
         assembly = Assembly(
             taxon_id=taxon_id,
             sample_id=long_read_specimen_sample_id,
             long_read_specimen_sample_id=long_read_specimen_sample_id,
             hic_specimen_sample_id=hic_specimen_sample_id,
+            hic_specimen_sample_ids=[str(sid) for sid in hic_specimen_sample_ids]
+            if hic_specimen_sample_ids
+            else None,
             data_types=data_types,
             version=version,
             tol_id=tol_id,
diff --git a/app/services/ncbi_taxonomy_service.py b/app/services/ncbi_taxonomy_service.py
index 998a9ce..32e303b 100644
--- a/app/services/ncbi_taxonomy_service.py
+++ b/app/services/ncbi_taxonomy_service.py
@@ -303,7 +303,7 @@ def extract_taxonomy_fields(report: dict[str, Any]) -> dict[str, Any]:
         full_lineage_future = executor.submit(process_parents, parents)
         mito_future = executor.submit(organelle_ref_lookup, parents, species_taxid, "Mitochondrion")
         ncbi_full_lineage = full_lineage_future.result()
-        mito_ref = mito_future.result()
+        mitohifi_reference_species = mito_future.result()
 
     return {
         "taxon_id": species_taxid,
@@ -318,7 +318,7 @@ def extract_taxonomy_fields(report: dict[str, Any]) -> dict[str, Any]:
         "ncbi_lineage": lineage,
         "ncbi_tax_string": lineage_to_string(lineage),
         "ncbi_full_lineage": ncbi_full_lineage,
-        "mito_ref": mito_ref,
+        "mitohifi_reference_species": mitohifi_reference_species,
     }
 
 
diff --git a/app/services/organism_service.py b/app/services/organism_service.py
index dc24c3e..91e618d 100644
--- a/app/services/organism_service.py
+++ b/app/services/organism_service.py
@@ -288,12 +288,7 @@ def update_organism(
         if not organism:
             return None
         non_bpa_fields = [
-            "ncbi_order",
-            "ncbi_family",
-            "busco_dataset_name",
-            "common_name",
-            "common_name_source",
-            "tax_string",
+            "scientific_name",
         ]
         new_bpa_json: Dict[str, Any] = organism.bpa_json or {}
         update_data = organism_in.model_dump(exclude_unset=True)
diff --git a/app/services/taxonomy_info_service.py b/app/services/taxonomy_info_service.py
index f40acd9..88cf9b0 100644
--- a/app/services/taxonomy_info_service.py
+++ b/app/services/taxonomy_info_service.py
@@ -160,13 +160,75 @@ def delete(self, db: Session, *, taxon_id: int) -> Optional[TaxonomyInfo]:
 
         if """
 
+    def _bulk_process_rows(
+        self,
+        db: Session,
+        *,
+        candidates: List[
+            tuple[int, Optional[TaxonomyInfoUpdate], Organism, Optional[TaxonomyInfo]]
+        ],
+        ncbi_by_taxon_id: Dict[int, Any],
+        apply_payload: bool = True,
+    ) -> tuple[int, int, int, List[str]]:
+        """Apply NCBI data (and optionally payload values) to a list of validated candidates.
+
+        Returns (created_count, updated_count, skipped_count, errors).
+        """
+        created_count = 0
+        updated_count = 0
+        skipped_count = 0
+        errors: List[str] = []
+
+        for taxon_id, row, organism, existing_ti in candidates:
+            try:
+                is_new = existing_ti is None
+                ti = existing_ti if existing_ti else TaxonomyInfo(taxon_id=taxon_id)
+                if is_new:
+                    db.add(ti)
+
+                mapped = ncbi_by_taxon_id.get(taxon_id)
+                if mapped:
+                    applied_fields = self._apply_ncbi_values(ti, mapped)
+                    ti.ncbi_last_synced_at = datetime.now(timezone.utc)
+                    logger.info(
+                        "NCBI taxonomy enrichment %s taxonomy_info for taxon_id=%s; applied_fields=%s",
+                        "created" if is_new else "updated",
+                        taxon_id,
+                        applied_fields,
+                    )
+                else:
+                    logger.warning(
+                        "NCBI enrichment returned no mapped taxonomy for taxon_id=%s",
+                        taxon_id,
+                    )
+
+                if apply_payload and row is not None:
+                    self._apply_payload_values(ti, row.model_dump(exclude_unset=True))
+
+                sync_organism_scientific_name(
+                    organism,
+                    ncbi_scientific_name=ti.ncbi_scientific_name,
+                )
+                db.add(organism)
+                db.commit()
+
+                if is_new:
+                    created_count += 1
+                else:
+                    updated_count += 1
+            except Exception as e:
+                errors.append(f"{taxon_id}: {str(e)}")
+                db.rollback()
+                skipped_count += 1
+
+        return created_count, updated_count, skipped_count, errors
+
     def bulk_import(
         self, db: Session, *, data: Dict[int, TaxonomyInfoUpdate]
     ) -> BulkImportResponse:
-        created_count = 0
         skipped_count = 0
         errors: List[str] = []
-        candidates: List[tuple[int, TaxonomyInfoUpdate, Organism]] = []
+        candidates: List[tuple[int, TaxonomyInfoUpdate, Organism, None]] = []
 
         for taxon_id, row in data.items():
             try:
@@ -182,61 +244,126 @@ def bulk_import(
                     )
                     skipped_count += 1
                     continue
-
             except Exception as e:
                 errors.append(f"{taxon_id}: {str(e)}")
                 db.rollback()
                 skipped_count += 1
                 continue
 
-            candidates.append((taxon_id, row, organism))
+            candidates.append((taxon_id, row, organism, None))
 
         scientific_names_by_taxon_id = {
             taxon_id: getattr(organism, "bpa_scientific_name", None)
-            for taxon_id, _, organism in candidates
+            for taxon_id, _, organism, _ in candidates
         }
         ncbi_by_taxon_id, unmapped = fetch_taxonomy_for_taxon_ids(scientific_names_by_taxon_id)
         if unmapped:
             logger.warning("NCBI bulk enrichment returned unmapped taxon_ids: %s", unmapped)
 
-        for taxon_id, row, organism in candidates:
-            try:
-                ti = TaxonomyInfo(taxon_id=taxon_id)
-                db.add(ti)
+        created_count, _, row_skipped, row_errors = self._bulk_process_rows(
+            db, candidates=candidates, ncbi_by_taxon_id=ncbi_by_taxon_id
+        )
+        skipped_count += row_skipped
+        errors.extend(row_errors)
 
-                mapped = ncbi_by_taxon_id.get(taxon_id)
-                if mapped:
-                    applied_fields = self._apply_ncbi_values(ti, mapped)
-                    ti.ncbi_last_synced_at = datetime.now(timezone.utc)
-                    logger.info(
-                        "NCBI taxonomy enrichment %s taxonomy_info for taxon_id=%s; applied_fields=%s",
-                        "created",
-                        taxon_id,
-                        applied_fields,
-                    )
-                else:
-                    logger.warning(
-                        "NCBI enrichment returned no mapped taxonomy during bulk import for taxon_id=%s",
-                        taxon_id,
-                    )
+        return BulkImportResponse(
+            created_count=created_count,
+            skipped_count=skipped_count,
+            message=f"TaxonomyInfo import complete. Created: {created_count}, Skipped: {skipped_count}",
+            errors=errors if errors else None,
+        )
 
-                self._apply_payload_values(ti, row.model_dump(exclude_unset=True))
-                sync_organism_scientific_name(
-                    organism,
-                    ncbi_scientific_name=ti.ncbi_scientific_name,
-                )
-                db.add(organism)
-                db.commit()
-                created_count += 1
+    def bulk_upsert(
+        self, db: Session, *, data: Dict[int, TaxonomyInfoUpdate]
+    ) -> BulkImportResponse:
+        skipped_count = 0
+        errors: List[str] = []
+        candidates: List[tuple[int, TaxonomyInfoUpdate, Organism, Optional[TaxonomyInfo]]] = []
+
+        for taxon_id, row in data.items():
+            try:
+                organism = db.query(Organism).filter(Organism.taxon_id == taxon_id).first()
+                if not organism:
+                    errors.append(f"{taxon_id}: organism with taxon_id {taxon_id} does not exist")
+                    skipped_count += 1
+                    continue
+                existing = db.query(TaxonomyInfo).filter(TaxonomyInfo.taxon_id == taxon_id).first()
             except Exception as e:
                 errors.append(f"{taxon_id}: {str(e)}")
                 db.rollback()
                 skipped_count += 1
+                continue
+
+            candidates.append((taxon_id, row, organism, existing))
+
+        scientific_names_by_taxon_id = {
+            taxon_id: getattr(organism, "bpa_scientific_name", None)
+            for taxon_id, _, organism, _ in candidates
+        }
+        ncbi_by_taxon_id, unmapped = fetch_taxonomy_for_taxon_ids(scientific_names_by_taxon_id)
+        if unmapped:
+            logger.warning("NCBI bulk enrichment returned unmapped taxon_ids: %s", unmapped)
+
+        created_count, updated_count, row_skipped, row_errors = self._bulk_process_rows(
+            db, candidates=candidates, ncbi_by_taxon_id=ncbi_by_taxon_id
+        )
+        skipped_count += row_skipped
+        errors.extend(row_errors)
 
         return BulkImportResponse(
             created_count=created_count,
+            updated_count=updated_count,
             skipped_count=skipped_count,
-            message=f"TaxonomyInfo import complete. Created: {created_count}, Skipped: {skipped_count}",
+            message=f"TaxonomyInfo upsert complete. Created: {created_count}, Updated: {updated_count}, Skipped: {skipped_count}",
+            errors=errors if errors else None,
+        )
+
+    def bulk_ncbi_refresh(self, db: Session, *, taxon_ids: List[int]) -> BulkImportResponse:
+        skipped_count = 0
+        errors: List[str] = []
+        candidates: List[tuple[int, None, Organism, TaxonomyInfo]] = []
+
+        for taxon_id in taxon_ids:
+            try:
+                organism = db.query(Organism).filter(Organism.taxon_id == taxon_id).first()
+                if not organism:
+                    errors.append(f"{taxon_id}: organism with taxon_id {taxon_id} does not exist")
+                    skipped_count += 1
+                    continue
+                existing = db.query(TaxonomyInfo).filter(TaxonomyInfo.taxon_id == taxon_id).first()
+                if not existing:
+                    errors.append(
+                        f"{taxon_id}: taxonomy_info for taxon_id {taxon_id} does not exist"
+                    )
+                    skipped_count += 1
+                    continue
+            except Exception as e:
+                errors.append(f"{taxon_id}: {str(e)}")
+                db.rollback()
+                skipped_count += 1
+                continue
+
+            candidates.append((taxon_id, None, organism, existing))
+
+        scientific_names_by_taxon_id = {
+            taxon_id: getattr(organism, "bpa_scientific_name", None)
+            for taxon_id, _, organism, _ in candidates
+        }
+        ncbi_by_taxon_id, unmapped = fetch_taxonomy_for_taxon_ids(scientific_names_by_taxon_id)
+        if unmapped:
+            logger.warning("NCBI bulk refresh returned unmapped taxon_ids: %s", unmapped)
+
+        _, updated_count, row_skipped, row_errors = self._bulk_process_rows(
+            db, candidates=candidates, ncbi_by_taxon_id=ncbi_by_taxon_id, apply_payload=False
+        )
+        skipped_count += row_skipped
+        errors.extend(row_errors)
+
+        return BulkImportResponse(
+            created_count=0,
+            updated_count=updated_count,
+            skipped_count=skipped_count,
+            message=f"NCBI taxonomy refresh complete. Updated: {updated_count}, Skipped: {skipped_count}",
             errors=errors if errors else None,
         )
 
diff --git a/schema.sql b/schema.sql
index ed60329..468a385 100644
--- a/schema.sql
+++ b/schema.sql
@@ -105,8 +105,6 @@ CREATE TABLE taxonomy_info (
     ncbi_tax_string TEXT,
     ncbi_full_lineage TEXT,
     ncbi_last_synced_at TIMESTAMPTZ,
-    mito_ref TEXT,
-    busco_dataset_name TEXT,
     busco_odb10_dataset_name TEXT,
     busco_odb12_dataset_name TEXT,
     find_plastid BOOLEAN,
@@ -114,7 +112,6 @@ CREATE TABLE taxonomy_info (
     mitochondrial_genetic_code_id INTEGER,
     mitohifi_reference_species TEXT,
     oatk_hmm_name TEXT,
-    defined_class TEXT,
     augustus_dataset_name TEXT,
     genetic_code_id INTEGER
 );
diff --git a/tests/unit/endpoints/test_endpoints_assemblies.py b/tests/unit/endpoints/test_endpoints_assemblies.py
index a86a764..92b94f3 100644
--- a/tests/unit/endpoints/test_endpoints_assemblies.py
+++ b/tests/unit/endpoints/test_endpoints_assemblies.py
@@ -808,7 +808,7 @@ def test_cancel_assembly_intent_success(monkeypatch):
         id=run_id,
         taxon_id=172942,
         long_read_specimen_sample_id=long_read_sample_id,
-        hic_specimen_sample_id=None,
+        hic_specimen_sample_ids=None,
         version=2,
         status="requested",
     )
diff --git a/tests/unit/endpoints/test_endpoints_taxonomy_info.py b/tests/unit/endpoints/test_endpoints_taxonomy_info.py
index 50fcb27..236881e 100644
--- a/tests/unit/endpoints/test_endpoints_taxonomy_info.py
+++ b/tests/unit/endpoints/test_endpoints_taxonomy_info.py
@@ -29,9 +29,7 @@ def _make_ti(taxon_id=1, **kwargs):
         "find_plastid": None,
         "hic_motif": None,
         "mitochondrial_genetic_code_id": None,
-        "mitohifi_reference_species": None,
         "oatk_hmm_name": None,
-        "defined_class": None,
         "augustus_dataset_name": None,
         "genetic_code_id": None,
     }
diff --git a/tests/unit/services/test_assembly_helper.py b/tests/unit/services/test_assembly_helper.py
index 5b9afa2..35627b6 100644
--- a/tests/unit/services/test_assembly_helper.py
+++ b/tests/unit/services/test_assembly_helper.py
@@ -31,11 +31,9 @@ def _make_mock_taxonomy_info():
         find_plastid=False,
         hic_motif="GATC",
         mitochondrial_genetic_code_id=1,
-        mitohifi_reference_species="test_species",
         oatk_hmm_name="test_hmm",
         augustus_dataset_name="test_augustus",
         genetic_code_id=1,
-        defined_class="test_class",
     )
 
 
@@ -753,11 +751,9 @@ def test_taxonomy_info_fields_included(self):
         assert result["find_plastid"] is False
         assert result["hic_motif"] == "GATC"
         assert result["mitochondrial_genetic_code_id"] == 1
-        assert result["mitohifi_reference_species"] == "test_species"
         assert result["oatk_hmm_name"] == "test_hmm"
         assert result["augustus_dataset_name"] == "test_augustus"
         assert result["genetic_code_id"] == 1
-        assert result["defined_class"] == "test_class"
 
     def test_sample_metadata_included_at_package_level(self):
         """Sample metadata (bpa_sample_id, specimen_id) appears at the package level."""
diff --git a/tests/unit/services/test_assembly_service.py b/tests/unit/services/test_assembly_service.py
index 7d55de0..5281677 100644
--- a/tests/unit/services/test_assembly_service.py
+++ b/tests/unit/services/test_assembly_service.py
@@ -283,7 +283,7 @@ def test_creates_assembly_with_requested_status(self, mock_db, assembly_service)
                 mock_db,
                 taxon_id=172942,
                 long_read_specimen_sample_id=LONG_READ_SAMPLE_ID,
-                hic_specimen_sample_id=None,
+                hic_specimen_sample_ids=None,
                 data_types="PACBIO_SMRT",
                 tol_id="tol-999",
                 project_id=None,
@@ -314,7 +314,7 @@ def capture_init(self, **kwargs):
                 mock_db,
                 taxon_id=172942,
                 long_read_specimen_sample_id=LONG_READ_SAMPLE_ID,
-                hic_specimen_sample_id=None,
+                hic_specimen_sample_ids=None,
                 data_types="PACBIO_SMRT",
                 tol_id=None,
                 project_id=None,
@@ -342,7 +342,7 @@ def capture_init(self, **kwargs):
                 mock_db,
                 taxon_id=172942,
                 long_read_specimen_sample_id=LONG_READ_SAMPLE_ID,
-                hic_specimen_sample_id=None,
+                hic_specimen_sample_ids=None,
                 data_types="PACBIO_SMRT",
                 tol_id=None,
                 project_id=None,
@@ -351,8 +351,8 @@ def capture_init(self, **kwargs):
 
         assert stored_manifests == [sample_manifest]
 
-    def test_hic_specimen_sample_id_stored(self, mock_db, assembly_service):
-        """hic_specimen_sample_id is stored on the Assembly."""
+    def test_hic_specimen_sample_ids_stored(self, mock_db, assembly_service):
+        """hic_specimen_sample_ids is stored as a JSON list; first entry sets hic_specimen_sample_id FK."""
         mock_query = Mock()
         mock_query.filter.return_value.scalar.return_value = None
         mock_db.query.return_value = mock_query
@@ -360,23 +360,29 @@ def test_hic_specimen_sample_id_stored(self, mock_db, assembly_service):
         mock_db.commit = Mock()
         mock_db.refresh = Mock()
 
-        hic_ids = []
+        captured = []
 
         def capture_init(self, **kwargs):
-            hic_ids.append(kwargs.get("hic_specimen_sample_id"))
+            captured.append(
+                {
+                    "hic_specimen_sample_id": kwargs.get("hic_specimen_sample_id"),
+                    "hic_specimen_sample_ids": kwargs.get("hic_specimen_sample_ids"),
+                }
+            )
 
         with patch.object(Assembly, "__init__", capture_init):
             assembly_service.create_from_intent(
                 mock_db,
                 taxon_id=172942,
                 long_read_specimen_sample_id=LONG_READ_SAMPLE_ID,
-                hic_specimen_sample_id=HIC_SAMPLE_ID,
+                hic_specimen_sample_ids=[HIC_SAMPLE_ID],
                 data_types="PACBIO_SMRT_HIC",
                 tol_id=None,
                 project_id=None,
             )
 
-        assert hic_ids == [HIC_SAMPLE_ID]
+        assert captured[0]["hic_specimen_sample_id"] == HIC_SAMPLE_ID
+        assert captured[0]["hic_specimen_sample_ids"] == [str(HIC_SAMPLE_ID)]
 
     def test_sample_id_set_to_long_read_specimen_sample_id(self, mock_db, assembly_service):
         """sample_id is set to long_read_specimen_sample_id for backward compatibility."""
@@ -397,7 +403,7 @@ def capture_init(self, **kwargs):
                 mock_db,
                 taxon_id=172942,
                 long_read_specimen_sample_id=LONG_READ_SAMPLE_ID,
-                hic_specimen_sample_id=None,
+                hic_specimen_sample_ids=None,
                 data_types="PACBIO_SMRT",
                 tol_id=None,
                 project_id=None,
diff --git a/tests/unit/services/test_taxonomy_info_service.py b/tests/unit/services/test_taxonomy_info_service.py
index 3439dc3..8af7d39 100644
--- a/tests/unit/services/test_taxonomy_info_service.py
+++ b/tests/unit/services/test_taxonomy_info_service.py
@@ -77,7 +77,7 @@ def test_populate_from_ncbi_lookup_creates_taxonomy_info(monkeypatch):
                     "ncbi_rank": "species",
                     "ncbi_scientific_name": "Penicillium test",
                     "ncbi_order": "Eurotiales",
-                    "mito_ref": "Penicillium chrysogenum",
+                    "mitohifi_reference_species": "Penicillium chrysogenum",
                 }
             },
             [],
@@ -97,7 +97,7 @@ def test_populate_from_ncbi_lookup_creates_taxonomy_info(monkeypatch):
     assert ti.ncbi_taxon_id == 5077
     assert ti.ncbi_rank == "species"
     assert ti.ncbi_order == "Eurotiales"
-    assert ti.mito_ref == "Penicillium chrysogenum"
+    assert ti.mitohifi_reference_species == "Penicillium chrysogenum"
     assert ti.ncbi_last_synced_at is not None
     assert organism.scientific_name == "Penicillium test"
     assert db.data[TaxonomyInfo][5077] is ti
@@ -176,14 +176,14 @@ def test_bulk_import_batches_ncbi_lookup_and_creates_taxonomy_info(monkeypatch):
                     "ncbi_taxon_id": 9612,
                     "ncbi_rank": "species",
                     "ncbi_scientific_name": "Canis lupus familiaris",
-                    "mito_ref": "Canis lupus familiaris",
+                    "mitohifi_reference_species": "Canis lupus familiaris",
                 },
                 9685: {
                     "taxon_id": 9685,
                     "ncbi_taxon_id": 9685,
                     "ncbi_rank": "species",
                     "ncbi_scientific_name": "Felis silvestris catus",
-                    "mito_ref": "Felis silvestris catus",
+                    "mitohifi_reference_species": "Felis silvestris catus",
                 },
             },
             [],
@@ -207,12 +207,12 @@ def test_bulk_import_batches_ncbi_lookup_and_creates_taxonomy_info(monkeypatch):
     saved_cat = db.data[TaxonomyInfo][9685]
     assert saved_dog.ncbi_taxon_id == 9612
     assert saved_dog.ncbi_rank == "species"
-    assert saved_dog.mito_ref == "Canis lupus familiaris"
+    assert saved_dog.mitohifi_reference_species == "Canis lupus familiaris"
     assert saved_dog.ncbi_last_synced_at is not None
     assert saved_dog.genetic_code_id == 2
     assert saved_cat.ncbi_taxon_id == 9685
     assert saved_cat.ncbi_rank == "species"
-    assert saved_cat.mito_ref == "Felis silvestris catus"
+    assert saved_cat.mitohifi_reference_species == "Felis silvestris catus"
     assert saved_cat.ncbi_last_synced_at is not None
     assert saved_cat.genetic_code_id == 1
     assert organisms[9612].scientific_name == "Canis lupus familiaris"