diff --git a/tests/test_transformer/test_nmdc_flattening_patterns.py b/tests/test_transformer/test_nmdc_flattening_patterns.py
new file mode 100644
index 0000000..89e9dad
--- /dev/null
+++ b/tests/test_transformer/test_nmdc_flattening_patterns.py
@@ -0,0 +1,533 @@
+"""Tests for NMDC-style flattening patterns using linkml-map expressions.
+
+These tests document how linkml-map can express the flattening operations
+currently performed by custom Python in
+``external-metadata-awareness/flatten_nmdc_collections.py``.
+
+Each NMDC AttributeValue type has a standard flattening pattern. The flat
+column naming conventions match those already used in the NMDC lakehouse
+(e.g. ``depth_has_numeric_value``, ``env_broad_scale_term_id``).
+
+Value types covered:
+- QuantityValue (has_numeric_value, has_unit, has_raw_value)
+- ControlledIdentifiedTermValue (term.id, term.name, has_raw_value)
+- GeolocationValue (latitude, longitude, has_raw_value)
+- TimestampValue (has_raw_value only)
+- PersonValue (name, email, orcid)
+
+See also:
+- PR #130 (basic QuantityValue flattening)
+- PR #131 (string-to-object construction, the inverse direction)
+"""
+
+import copy
+
+from linkml_runtime import SchemaView
+
+from linkml_map.transformer.object_transformer import ObjectTransformer
+
+# ---------------------------------------------------------------------------
+# Source schema: NMDC-style with nested AttributeValue types
+# ---------------------------------------------------------------------------
+
+SOURCE_SCHEMA = """\
+id: https://example.org/nmdc-source
+name: nmdc_source
+prefixes:
+  linkml: https://w3id.org/linkml/
+  nmdc: https://w3id.org/nmdc/
+imports:
+  - linkml:types
+
+classes:
+  Biosample:
+    tree_root: true
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+      depth:
+        range: QuantityValue
+        inlined: true
+      temp:
+        range: QuantityValue
+        inlined: true
+      env_broad_scale:
+        range: ControlledIdentifiedTermValue
+        inlined: true
+      env_local_scale:
+        range: ControlledIdentifiedTermValue
+        inlined: true
+      env_medium:
+        range: ControlledIdentifiedTermValue
+        inlined: true
+      lat_lon:
+        range: GeolocationValue
+        inlined: true
+      collection_date:
+        range: TimestampValue
+        inlined: true
+      associated_studies:
+        range: string
+        multivalued: true
+      ecosystem:
+        range: string
+      ecosystem_category:
+        range: string
+
+  QuantityValue:
+    attributes:
+      has_raw_value:
+        range: string
+      has_numeric_value:
+        range: float
+      has_unit:
+        range: string
+      has_minimum_numeric_value:
+        range: float
+      has_maximum_numeric_value:
+        range: float
+
+  OntologyClass:
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+
+  ControlledIdentifiedTermValue:
+    attributes:
+      has_raw_value:
+        range: string
+      term:
+        range: OntologyClass
+        inlined: true
+        required: true
+
+  GeolocationValue:
+    attributes:
+      has_raw_value:
+        range: string
+      latitude:
+        range: float
+        required: true
+      longitude:
+        range: float
+        required: true
+
+  TimestampValue:
+    attributes:
+      has_raw_value:
+        range: string
+
+  Study:
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+      principal_investigator:
+        range: PersonValue
+        inlined: true
+      ecosystem:
+        range: string
+
+  PersonValue:
+    attributes:
+      name:
+        range: string
+      email:
+        range: string
+      orcid:
+        range: string
+"""
+
+# ---------------------------------------------------------------------------
+# Target schema: flat lakehouse columns
+# ---------------------------------------------------------------------------
+
+TARGET_SCHEMA = """\
+id: https://example.org/nmdc-flat
+name: nmdc_flat
+prefixes:
+  linkml: https://w3id.org/linkml/
+imports:
+  - linkml:types
+
+classes:
+  FlatBiosample:
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+      # QuantityValue flattened
+      depth_has_numeric_value:
+        range: float
+      depth_has_unit:
+        range: string
+      depth_has_raw_value:
+        range: string
+      depth_has_minimum_numeric_value:
+        range: float
+      depth_has_maximum_numeric_value:
+        range: float
+      temp_has_numeric_value:
+        range: float
+      temp_has_unit:
+        range: string
+      # ControlledIdentifiedTermValue flattened
+      env_broad_scale_term_id:
+        range: string
+      env_broad_scale_term_name:
+        range: string
+      env_broad_scale_has_raw_value:
+        range: string
+      env_local_scale_term_id:
+        range: string
+      env_local_scale_term_name:
+        range: string
+      env_medium_term_id:
+        range: string
+      env_medium_term_name:
+        range: string
+      # GeolocationValue flattened
+      lat_lon_latitude:
+        range: float
+      lat_lon_longitude:
+        range: float
+      lat_lon_has_raw_value:
+        range: string
+      # TimestampValue flattened
+      collection_date_has_raw_value:
+        range: string
+      # Scalar passthrough
+      ecosystem:
+        range: string
+      ecosystem_category:
+        range: string
+
+  FlatStudy:
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+      pi_name:
+        range: string
+      pi_email:
+        range: string
+      pi_orcid:
+        range: string
+      ecosystem:
+        range: string
+"""
+
+# ---------------------------------------------------------------------------
+# Transformation specs
+# ---------------------------------------------------------------------------
+
+BIOSAMPLE_FLATTEN_SPEC = {
+    "class_derivations": {
+        "FlatBiosample": {
+            "populated_from": "Biosample",
+            "slot_derivations": {
+                "id": {},
+                "name": {},
+                # QuantityValue: depth (full 5-field pattern)
+                "depth_has_numeric_value": {"expr": "depth.has_numeric_value"},
+                "depth_has_unit": {"expr": "depth.has_unit"},
+                "depth_has_raw_value": {"expr": "depth.has_raw_value"},
+                "depth_has_minimum_numeric_value": {"expr": "depth.has_minimum_numeric_value"},
+                "depth_has_maximum_numeric_value": {"expr": "depth.has_maximum_numeric_value"},
+                # QuantityValue: temp (partial)
+                "temp_has_numeric_value": {"expr": "temp.has_numeric_value"},
+                "temp_has_unit": {"expr": "temp.has_unit"},
+                # ControlledIdentifiedTermValue: env triad (2-level nesting)
+                "env_broad_scale_term_id": {"expr": "env_broad_scale.term.id"},
+                "env_broad_scale_term_name": {"expr": "env_broad_scale.term.name"},
+                "env_broad_scale_has_raw_value": {"expr": "env_broad_scale.has_raw_value"},
+                "env_local_scale_term_id": {"expr": "env_local_scale.term.id"},
+                "env_local_scale_term_name": {"expr": "env_local_scale.term.name"},
+                "env_medium_term_id": {"expr": "env_medium.term.id"},
+                "env_medium_term_name": {"expr": "env_medium.term.name"},
+                # GeolocationValue: lat_lon
+                "lat_lon_latitude": {"expr": "lat_lon.latitude"},
+                "lat_lon_longitude": {"expr": "lat_lon.longitude"},
+                "lat_lon_has_raw_value": {"expr": "lat_lon.has_raw_value"},
+                # TimestampValue: collection_date
+                "collection_date_has_raw_value": {"expr": "collection_date.has_raw_value"},
+                # Scalar passthrough
+                "ecosystem": {},
+                "ecosystem_category": {},
+            },
+        }
+    }
+}
+
+STUDY_FLATTEN_SPEC = {
+    "class_derivations": {
+        "FlatStudy": {
+            "populated_from": "Study",
+            "slot_derivations": {
+                "id": {},
+                "name": {},
+                "pi_name": {"expr": "principal_investigator.name"},
+                "pi_email": {"expr": "principal_investigator.email"},
+                "pi_orcid": {"expr": "principal_investigator.orcid"},
+                "ecosystem": {},
+            },
+        }
+    }
+}
+
+
+# ---------------------------------------------------------------------------
+# Helpers
+# ---------------------------------------------------------------------------
+
+def _make_transformer(spec):
+    """Build an ObjectTransformer wired to the NMDC source/target schemas."""
+    tr = ObjectTransformer()
+    tr.source_schemaview = SchemaView(SOURCE_SCHEMA)
+    tr.target_schemaview = SchemaView(TARGET_SCHEMA)
+    tr.create_transformer_specification(copy.deepcopy(spec))
+    return tr
+
+
+# ---------------------------------------------------------------------------
+# Test data fixtures
+# ---------------------------------------------------------------------------
+
+COMPLETE_BIOSAMPLE = {
+    "id": "nmdc:bsm-12-abc123",
+    "name": "Terrestrial soil  - Hopland",
+    "depth": {
+        "has_raw_value": "0.5 to 1.0 m",
+        "has_numeric_value": 0.75,
+        "has_unit": "m",
+        "has_minimum_numeric_value": 0.5,
+        "has_maximum_numeric_value": 1.0,
+    },
+    "temp": {
+        "has_numeric_value": 18.3,
+        "has_unit": "degree Celsius",
+    },
+    "env_broad_scale": {
+        "has_raw_value": "forest biome [ENVO:01000174]",
+        "term": {"id": "ENVO:01000174", "name": "forest biome"},
+    },
+    "env_local_scale": {
+        "has_raw_value": "temperate mixed forest [ENVO:01000216]",
+        "term": {"id": "ENVO:01000216", "name": "temperate mixed forest"},
+    },
+    "env_medium": {
+        "has_raw_value": "bulk soil [ENVO:00002259]",
+        "term": {"id": "ENVO:00002259", "name": "bulk soil"},
+    },
+    "lat_lon": {
+        "has_raw_value": "38.993 -123.067",
+        "latitude": 38.993,
+        "longitude": -123.067,
+    },
+    "collection_date": {
+        "has_raw_value": "2021-06-15",
+    },
+    "associated_studies": ["nmdc:sty-11-study01"],
+    "ecosystem": "Environmental",
+    "ecosystem_category": "Terrestrial",
+}
+
+SPARSE_BIOSAMPLE = {
+    "id": "nmdc:bsm-12-sparse99",
+    "name": "Minimal sample",
+    "depth": None,
+    "temp": None,
+    "env_broad_scale": {
+        "has_raw_value": "aquatic biome [ENVO:00002030]",
+        "term": {"id": "ENVO:00002030", "name": "aquatic biome"},
+    },
+    "env_local_scale": None,
+    "env_medium": None,
+    "lat_lon": {
+        "latitude": 45.0,
+        "longitude": -120.0,
+    },
+    "collection_date": None,
+    "ecosystem": None,
+    "ecosystem_category": None,
+}
+
+
+# ===================================================================
+# Biosample flattening tests
+# ===================================================================
+
+
+class TestBiosampleFlatten:
+    """Tests for flattening Biosample to FlatBiosample."""
+
+    def test_complete_biosample(self):
+        """All fields populated: every flat column gets a value."""
+        tr = _make_transformer(BIOSAMPLE_FLATTEN_SPEC)
+        result = tr.map_object(COMPLETE_BIOSAMPLE, source_type="Biosample")
+
+        # Identity
+        assert result["id"] == "nmdc:bsm-12-abc123"
+        assert result["name"] == "Terrestrial soil  - Hopland"
+
+        # QuantityValue: depth (full)
+        assert result["depth_has_numeric_value"] == 0.75
+        assert result["depth_has_unit"] == "m"
+        assert result["depth_has_raw_value"] == "0.5 to 1.0 m"
+        assert result["depth_has_minimum_numeric_value"] == 0.5
+        assert result["depth_has_maximum_numeric_value"] == 1.0
+
+        # QuantityValue: temp
+        assert result["temp_has_numeric_value"] == 18.3
+        assert result["temp_has_unit"] == "degree Celsius"
+
+        # ControlledIdentifiedTermValue: env triad (2-level nesting)
+        assert result["env_broad_scale_term_id"] == "ENVO:01000174"
+        assert result["env_broad_scale_term_name"] == "forest biome"
+        assert result["env_broad_scale_has_raw_value"] == "forest biome [ENVO:01000174]"
+        assert result["env_local_scale_term_id"] == "ENVO:01000216"
+        assert result["env_local_scale_term_name"] == "temperate mixed forest"
+        assert result["env_medium_term_id"] == "ENVO:00002259"
+        assert result["env_medium_term_name"] == "bulk soil"
+
+        # GeolocationValue
+        assert result["lat_lon_latitude"] == 38.993
+        assert result["lat_lon_longitude"] == -123.067
+        assert result["lat_lon_has_raw_value"] == "38.993 -123.067"
+
+        # TimestampValue
+        assert result["collection_date_has_raw_value"] == "2021-06-15"
+
+        # Scalar passthrough
+        assert result["ecosystem"] == "Environmental"
+        assert result["ecosystem_category"] == "Terrestrial"
+
+    def test_sparse_biosample_null_propagation(self):
+        """Null QuantityValue/ControlledTermValue propagates to None flat columns."""
+        tr = _make_transformer(BIOSAMPLE_FLATTEN_SPEC)
+        result = tr.map_object(SPARSE_BIOSAMPLE, source_type="Biosample")
+
+        assert result["id"] == "nmdc:bsm-12-sparse99"
+
+        # Null depth → all depth columns None
+        assert result["depth_has_numeric_value"] is None
+        assert result["depth_has_unit"] is None
+        assert result["depth_has_raw_value"] is None
+        assert result["depth_has_minimum_numeric_value"] is None
+        assert result["depth_has_maximum_numeric_value"] is None
+
+        # Null temp
+        assert result["temp_has_numeric_value"] is None
+        assert result["temp_has_unit"] is None
+
+        # env_broad_scale populated, others null
+        assert result["env_broad_scale_term_id"] == "ENVO:00002030"
+        assert result["env_broad_scale_term_name"] == "aquatic biome"
+        assert result["env_local_scale_term_id"] is None
+        assert result["env_local_scale_term_name"] is None
+        assert result["env_medium_term_id"] is None
+        assert result["env_medium_term_name"] is None
+
+        # lat_lon populated but missing has_raw_value
+        assert result["lat_lon_latitude"] == 45.0
+        assert result["lat_lon_longitude"] == -120.0
+        assert result["lat_lon_has_raw_value"] is None
+
+        # Null collection_date
+        assert result["collection_date_has_raw_value"] is None
+
+    def test_partial_quantity_value(self):
+        """QuantityValue with some fields missing: populated fields come through."""
+        tr = _make_transformer(BIOSAMPLE_FLATTEN_SPEC)
+        source = {
+            "id": "nmdc:bsm-12-partial",
+            "name": "Partial QV",
+            "depth": {"has_numeric_value": 3.0},  # no unit, no raw_value
+            "temp": {"has_unit": "degree Celsius"},  # no numeric_value
+            "env_broad_scale": None,
+            "env_local_scale": None,
+            "env_medium": None,
+            "lat_lon": {"latitude": 0.0, "longitude": 0.0},
+            "collection_date": None,
+        }
+        result = tr.map_object(source, source_type="Biosample")
+
+        assert result["depth_has_numeric_value"] == 3.0
+        assert result["depth_has_unit"] is None
+        assert result["depth_has_raw_value"] is None
+        assert result["temp_has_numeric_value"] is None
+        assert result["temp_has_unit"] == "degree Celsius"
+
+
+# ===================================================================
+# Study flattening tests (PersonValue)
+# ===================================================================
+
+
+class TestStudyFlatten:
+    """Tests for flattening Study with nested PersonValue."""
+
+    def test_complete_study(self):
+        """Study with PI: PersonValue fields are extracted."""
+        tr = _make_transformer(STUDY_FLATTEN_SPEC)
+        source = {
+            "id": "nmdc:sty-11-study01",
+            "name": "Hopland Research Station Soil Metagenome",
+            "principal_investigator": {
+                "name": "Elaine Faustman",
+                "email": "faustman@example.org",
+                "orcid": "0000-0001-2345-6789",
+            },
+            "ecosystem": "Environmental",
+        }
+        result = tr.map_object(source, source_type="Study")
+
+        assert result["id"] == "nmdc:sty-11-study01"
+        assert result["name"] == "Hopland Research Station Soil Metagenome"
+        assert result["pi_name"] == "Elaine Faustman"
+        assert result["pi_email"] == "faustman@example.org"
+        assert result["pi_orcid"] == "0000-0001-2345-6789"
+        assert result["ecosystem"] == "Environmental"
+
+    def test_study_no_pi(self):
+        """Study with no PI: PersonValue fields are None."""
+        tr = _make_transformer(STUDY_FLATTEN_SPEC)
+        source = {
+            "id": "nmdc:sty-11-nopi",
+            "name": "Study Without PI",
+            "principal_investigator": None,
+            "ecosystem": None,
+        }
+        result = tr.map_object(source, source_type="Study")
+
+        assert result["pi_name"] is None
+        assert result["pi_email"] is None
+        assert result["pi_orcid"] is None
+
+    def test_study_partial_pi(self):
+        """PI with only name, no email/orcid."""
+        tr = _make_transformer(STUDY_FLATTEN_SPEC)
+        source = {
+            "id": "nmdc:sty-11-partpi",
+            "name": "Partial PI Study",
+            "principal_investigator": {
+                "name": "Jane Doe",
+            },
+        }
+        result = tr.map_object(source, source_type="Study")
+
+        assert result["pi_name"] == "Jane Doe"
+        assert result["pi_email"] is None
+        assert result["pi_orcid"] is None
diff --git a/tests/test_transformer/test_nmdc_unified_activities.py b/tests/test_transformer/test_nmdc_unified_activities.py
new file mode 100644
index 0000000..041e533
--- /dev/null
+++ b/tests/test_transformer/test_nmdc_unified_activities.py
@@ -0,0 +1,321 @@
+"""Tests for merging multiple NMDC workflow activity types into a unified flat table.
+
+Demonstrates the multiple-class_derivations-for-same-target feature (#118):
+several source classes (MetagenomeSequencing, ReadQcAnalysis, MetagenomeAssembly)
+are each mapped to the same ``FlatActivity`` target class, producing a single
+unified stream of activity records with a ``activity_type`` discriminator column.
+
+This is the pattern needed for the NMDC lakehouse "all activities" table,
+replacing custom Python union queries across N separate MongoDB collections.
+
+See also: PR #118 (Allow multiple class_derivations for the same target class)
+"""
+
+import copy
+
+import pytest
+from linkml_runtime import SchemaView
+
+from linkml_map.transformer.object_transformer import ObjectTransformer
+
+# ---------------------------------------------------------------------------
+# Source schema: multiple workflow activity classes
+# ---------------------------------------------------------------------------
+
+SOURCE_SCHEMA = """\
+id: https://example.org/nmdc-activities-source
+name: nmdc_activities_source
+prefixes:
+  linkml: https://w3id.org/linkml/
+imports:
+  - linkml:types
+
+classes:
+  WorkflowExecution:
+    abstract: true
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+      was_informed_by:
+        range: string
+      started_at_time:
+        range: string
+      ended_at_time:
+        range: string
+      execution_resource:
+        range: string
+      git_url:
+        range: string
+
+  MetagenomeSequencing:
+    is_a: WorkflowExecution
+    attributes:
+      analyte_category:
+        range: string
+      instrument_used:
+        range: string
+
+  ReadQcAnalysis:
+    is_a: WorkflowExecution
+    attributes:
+      input_read_count:
+        range: integer
+      output_read_count:
+        range: integer
+      input_base_count:
+        range: integer
+      output_base_count:
+        range: integer
+
+  MetagenomeAssembly:
+    is_a: WorkflowExecution
+    attributes:
+      scaffolds:
+        range: integer
+      scaf_bp:
+        range: integer
+      contigs:
+        range: integer
+      gap_pct:
+        range: float
+"""
+
+# ---------------------------------------------------------------------------
+# Target schema: unified flat activity table
+# ---------------------------------------------------------------------------
+
+TARGET_SCHEMA = """\
+id: https://example.org/nmdc-activities-flat
+name: nmdc_activities_flat
+prefixes:
+  linkml: https://w3id.org/linkml/
+imports:
+  - linkml:types
+
+classes:
+  FlatActivity:
+    attributes:
+      activity_id:
+        identifier: true
+        range: string
+      activity_type:
+        range: string
+      name:
+        range: string
+      was_informed_by:
+        range: string
+      started_at:
+        range: string
+      ended_at:
+        range: string
+      execution_resource:
+        range: string
+"""
+
+# ---------------------------------------------------------------------------
+# Transformation spec: three class_derivations for the same target
+# ---------------------------------------------------------------------------
+
+# Note: with #118, multiple class_derivations entries with the same target
+# class name are stored as a list (inlined_as_list). Each derivation pulls
+# from a different source class.
+
+UNIFIED_ACTIVITY_SPEC = {
+    "class_derivations": [
+        {
+            "name": "FlatActivity",
+            "populated_from": "MetagenomeSequencing",
+            "slot_derivations": {
+                "activity_id": {"populated_from": "id"},
+                "activity_type": {"value": "MetagenomeSequencing"},
+                "name": {},
+                "was_informed_by": {},
+                "started_at": {"populated_from": "started_at_time"},
+                "ended_at": {"populated_from": "ended_at_time"},
+                "execution_resource": {},
+            },
+        },
+        {
+            "name": "FlatActivity",
+            "populated_from": "ReadQcAnalysis",
+            "slot_derivations": {
+                "activity_id": {"populated_from": "id"},
+                "activity_type": {"value": "ReadQcAnalysis"},
+                "name": {},
+                "was_informed_by": {},
+                "started_at": {"populated_from": "started_at_time"},
+                "ended_at": {"populated_from": "ended_at_time"},
+                "execution_resource": {},
+            },
+        },
+        {
+            "name": "FlatActivity",
+            "populated_from": "MetagenomeAssembly",
+            "slot_derivations": {
+                "activity_id": {"populated_from": "id"},
+                "activity_type": {"value": "MetagenomeAssembly"},
+                "name": {},
+                "was_informed_by": {},
+                "started_at": {"populated_from": "started_at_time"},
+                "ended_at": {"populated_from": "ended_at_time"},
+                "execution_resource": {},
+            },
+        },
+    ]
+}
+
+
+# ---------------------------------------------------------------------------
+# Test data
+# ---------------------------------------------------------------------------
+
+SEQUENCING_RECORD = {
+    "id": "nmdc:dgns-11-seq001",
+    "name": "Sequencing of Hopland soil",
+    "was_informed_by": "nmdc:omprc-11-study01",
+    "started_at_time": "2021-06-15T08:00:00",
+    "ended_at_time": "2021-06-15T16:00:00",
+    "execution_resource": "JGI",
+    "git_url": "https://github.com/example/pipeline",
+    "analyte_category": "metagenome",
+    "instrument_used": "Illumina NovaSeq",
+}
+
+QC_RECORD = {
+    "id": "nmdc:wfrqc-11-qc001",
+    "name": "Read QC for Hopland",
+    "was_informed_by": "nmdc:dgns-11-seq001",
+    "started_at_time": "2021-06-16T00:00:00",
+    "ended_at_time": "2021-06-16T02:00:00",
+    "execution_resource": "NERSC - Perlmutter",
+    "git_url": "https://github.com/example/rqc",
+    "input_read_count": 50000000,
+    "output_read_count": 48000000,
+    "input_base_count": 7500000000,
+    "output_base_count": 7200000000,
+}
+
+ASSEMBLY_RECORD = {
+    "id": "nmdc:wfmgas-11-asm001",
+    "name": "Assembly of Hopland metagenome",
+    "was_informed_by": "nmdc:wfrqc-11-qc001",
+    "started_at_time": "2021-06-17T00:00:00",
+    "ended_at_time": "2021-06-17T12:00:00",
+    "execution_resource": "NERSC - Perlmutter",
+    "git_url": "https://github.com/example/assembly",
+    "scaffolds": 15000,
+    "scaf_bp": 120000000,
+    "contigs": 18000,
+    "gap_pct": 1.5,
+}
+
+
+# ---------------------------------------------------------------------------
+# Helper
+# ---------------------------------------------------------------------------
+
+def _make_transformer(spec):
+    tr = ObjectTransformer()
+    tr.source_schemaview = SchemaView(SOURCE_SCHEMA)
+    tr.target_schemaview = SchemaView(TARGET_SCHEMA)
+    tr.create_transformer_specification(copy.deepcopy(spec))
+    return tr
+
+
+# ---------------------------------------------------------------------------
+# Tests
+# ---------------------------------------------------------------------------
+
+
+def test_sequencing_to_flat_activity():
+    """MetagenomeSequencing maps to FlatActivity with correct type discriminator."""
+    tr = _make_transformer(UNIFIED_ACTIVITY_SPEC)
+    result = tr.map_object(SEQUENCING_RECORD, source_type="MetagenomeSequencing")
+
+    assert result["activity_id"] == "nmdc:dgns-11-seq001"
+    assert result["activity_type"] == "MetagenomeSequencing"
+    assert result["name"] == "Sequencing of Hopland soil"
+    assert result["was_informed_by"] == "nmdc:omprc-11-study01"
+    assert result["started_at"] == "2021-06-15T08:00:00"
+    assert result["ended_at"] == "2021-06-15T16:00:00"
+    assert result["execution_resource"] == "JGI"
+
+
+def test_qc_to_flat_activity():
+    """ReadQcAnalysis maps to FlatActivity; QC-specific fields are not carried."""
+    tr = _make_transformer(UNIFIED_ACTIVITY_SPEC)
+    result = tr.map_object(QC_RECORD, source_type="ReadQcAnalysis")
+
+    assert result["activity_id"] == "nmdc:wfrqc-11-qc001"
+    assert result["activity_type"] == "ReadQcAnalysis"
+    assert result["name"] == "Read QC for Hopland"
+    assert result["started_at"] == "2021-06-16T00:00:00"
+    assert result["execution_resource"] == "NERSC - Perlmutter"
+    # QC-specific fields are not in FlatActivity target schema
+    assert "input_read_count" not in result
+    assert "output_read_count" not in result
+
+
+def test_assembly_to_flat_activity():
+    """MetagenomeAssembly maps to FlatActivity; assembly stats not carried."""
+    tr = _make_transformer(UNIFIED_ACTIVITY_SPEC)
+    result = tr.map_object(ASSEMBLY_RECORD, source_type="MetagenomeAssembly")
+
+    assert result["activity_id"] == "nmdc:wfmgas-11-asm001"
+    assert result["activity_type"] == "MetagenomeAssembly"
+    assert result["name"] == "Assembly of Hopland metagenome"
+    assert result["was_informed_by"] == "nmdc:wfrqc-11-qc001"
+    assert result["started_at"] == "2021-06-17T00:00:00"
+    assert result["ended_at"] == "2021-06-17T12:00:00"
+    # Assembly-specific fields are not in target
+    assert "scaffolds" not in result
+    assert "gap_pct" not in result
+
+
+def test_all_types_produce_distinct_records():
+    """Each activity type produces a distinct record with correct type tag."""
+    tr = _make_transformer(UNIFIED_ACTIVITY_SPEC)
+
+    results = [
+        tr.map_object(SEQUENCING_RECORD, source_type="MetagenomeSequencing"),
+        tr.map_object(QC_RECORD, source_type="ReadQcAnalysis"),
+        tr.map_object(ASSEMBLY_RECORD, source_type="MetagenomeAssembly"),
+    ]
+
+    types = [r["activity_type"] for r in results]
+    assert types == ["MetagenomeSequencing", "ReadQcAnalysis", "MetagenomeAssembly"]
+
+    ids = [r["activity_id"] for r in results]
+    assert len(set(ids)) == 3  # all distinct
+
+
+def test_slot_rename_started_at_time():
+    """started_at_time is renamed to started_at via populated_from."""
+    tr = _make_transformer(UNIFIED_ACTIVITY_SPEC)
+    result = tr.map_object(QC_RECORD, source_type="ReadQcAnalysis")
+
+    assert "started_at" in result
+    assert "started_at_time" not in result
+    assert result["started_at"] == "2021-06-16T00:00:00"
+
+
+def test_missing_optional_field():
+    """Optional fields missing from source produce None in target."""
+    tr = _make_transformer(UNIFIED_ACTIVITY_SPEC)
+    sparse = {
+        "id": "nmdc:wfrqc-11-sparse",
+        "name": "Sparse QC",
+        "was_informed_by": "nmdc:dgns-11-seq001",
+        "started_at_time": "2021-01-01T00:00:00",
+        # ended_at_time, execution_resource, git_url all missing
+    }
+    result = tr.map_object(sparse, source_type="ReadQcAnalysis")
+
+    assert result["activity_id"] == "nmdc:wfrqc-11-sparse"
+    assert result["activity_type"] == "ReadQcAnalysis"
+    assert result["started_at"] == "2021-01-01T00:00:00"
+    assert result.get("ended_at") is None
+    assert result.get("execution_resource") is None
diff --git a/tests/test_transformer/test_nmdc_uuid5_ids.py b/tests/test_transformer/test_nmdc_uuid5_ids.py
new file mode 100644
index 0000000..21ac707
--- /dev/null
+++ b/tests/test_transformer/test_nmdc_uuid5_ids.py
@@ -0,0 +1,145 @@
+"""Tests for deterministic ID generation using uuid5() in transformation expressions.
+
+NMDC uses typed identifiers with a specific pattern: ``nmdc:{typecode}-{shoulder}-{blade}``.
+For lakehouse ETL where the same source data is re-processed, deterministic IDs ensure
+idempotent results. The ``uuid5()`` function in linkml-map (#117) enables this.
+
+``uuid5(namespace_url, name_string)`` produces a deterministic UUID from a namespace
+and a name. By using a fixed namespace and composing the name from source fields,
+the same input always produces the same output ID.
+"""
+
+import copy
+import uuid
+
+from linkml_runtime import SchemaView
+
+from linkml_map.transformer.object_transformer import ObjectTransformer
+
+SOURCE_SCHEMA = """\
+id: https://example.org/ext-source
+name: ext_source
+prefixes:
+  linkml: https://w3id.org/linkml/
+imports:
+  - linkml:types
+classes:
+  SourceRecord:
+    attributes:
+      source_db:
+        range: string
+      source_id:
+        identifier: true
+        range: string
+      sample_name:
+        range: string
+"""
+
+TARGET_SCHEMA = """\
+id: https://example.org/nmdc-target
+name: nmdc_target
+prefixes:
+  linkml: https://w3id.org/linkml/
+imports:
+  - linkml:types
+classes:
+  Biosample:
+    attributes:
+      id:
+        identifier: true
+        range: string
+      name:
+        range: string
+      description:
+        range: string
+"""
+
+TRANSFORM_SPEC = {
+    "class_derivations": {
+        "Biosample": {
+            "populated_from": "SourceRecord",
+            "slot_derivations": {
+                "id": {
+                    "expr": "'nmdc:bsm-' + uuid5('https://w3id.org/nmdc/', {source_db} + ':' + {source_id})",
+                },
+                "name": {"populated_from": "sample_name"},
+                "description": {
+                    "expr": "'Sample from ' + {source_db} + ' study ' + {source_id}",
+                },
+            },
+        }
+    }
+}
+
+
+def _make_transformer():
+    tr = ObjectTransformer()
+    tr.source_schemaview = SchemaView(SOURCE_SCHEMA)
+    tr.target_schemaview = SchemaView(TARGET_SCHEMA)
+    tr.create_transformer_specification(copy.deepcopy(TRANSFORM_SPEC))
+    return tr
+
+
+def test_uuid5_deterministic():
+    """Same input produces the same ID every time."""
+    tr = _make_transformer()
+    source = {"source_db": "GOLD", "source_id": "Gb0001234", "sample_name": "Forest soil"}
+
+    result1 = tr.map_object(source, source_type="SourceRecord")
+    result2 = tr.map_object(source, source_type="SourceRecord")
+
+    assert result1["id"] == result2["id"]
+    assert result1["id"].startswith("nmdc:bsm-")
+
+
+def test_uuid5_different_inputs_different_ids():
+    """Different source records produce different IDs."""
+    tr = _make_transformer()
+
+    result_a = tr.map_object(
+        {"source_db": "GOLD", "source_id": "Gb0001234", "sample_name": "A"},
+        source_type="SourceRecord",
+    )
+    result_b = tr.map_object(
+        {"source_db": "GOLD", "source_id": "Gb0005678", "sample_name": "B"},
+        source_type="SourceRecord",
+    )
+
+    assert result_a["id"] != result_b["id"]
+
+
+def test_uuid5_matches_python_stdlib():
+    """The uuid5 output matches Python's uuid.uuid5 for the same inputs."""
+    tr = _make_transformer()
+    source = {"source_db": "GOLD", "source_id": "Gb0001234", "sample_name": "Test"}
+
+    result = tr.map_object(source, source_type="SourceRecord")
+
+    # Compute expected UUID using stdlib
+    namespace = uuid.uuid5(uuid.NAMESPACE_URL, "https://w3id.org/nmdc/")
+    expected_uuid = str(uuid.uuid5(namespace, "GOLD:Gb0001234"))
+    expected_id = f"nmdc:bsm-{expected_uuid}"
+
+    assert result["id"] == expected_id
+
+
+def test_uuid5_with_string_concatenation():
+    """Description field uses string concatenation with source fields."""
+    tr = _make_transformer()
+    source = {"source_db": "GOLD", "source_id": "Gb0001234", "sample_name": "Forest soil"}
+
+    result = tr.map_object(source, source_type="SourceRecord")
+
+    assert result["name"] == "Forest soil"
+    assert result["description"] == "Sample from GOLD study Gb0001234"
+
+
+def test_uuid5_null_source_field():
+    """When a source field used in uuid5 is None, null propagation kicks in."""
+    tr = _make_transformer()
+    source = {"source_db": None, "source_id": "Gb0001234", "sample_name": "Test"}
+
+    result = tr.map_object(source, source_type="SourceRecord")
+
+    # {source_db} is None → null propagation → entire expr is None
+    assert result["id"] is None