lamindb 1.0.4__py3-none-any.whl → 1.1.0__py3-none-any.whl

lamindb/curators/__init__.py

@@ -1,21 +1,52 @@
+"""Curators.
+
+.. autosummary::
+   :toctree: .
+
+   Curator
+   DataFrameCurator
+   AnnDataCurator
+
+"""
+
 from __future__ import annotations
 
 import copy
-import warnings
+import random
+import re
+from importlib import resources
 from itertools import chain
-from typing import TYPE_CHECKING
+from typing import TYPE_CHECKING, Any, Literal
 
 import anndata as ad
 import lamindb_setup as ln_setup
 import pandas as pd
+import pandera
 import pyarrow as pa
 from lamin_utils import colors, logger
+from lamindb_setup.core import deprecated, upath
 from lamindb_setup.core._docs import doc_args
 from lamindb_setup.core.upath import UPath
 
+from lamindb.core.storage._backed_access import backed_access
+
+from ._cellxgene_schemas import _read_schema_versions
+
+if TYPE_CHECKING:
+    from anndata import AnnData
+    from lamindb_setup.core.types import UPathStr
+
+    from lamindb.base.types import FieldAttr
+    from lamindb.models import Record
+from lamindb._feature import parse_dtype, parse_dtype_single_cat
 from lamindb.base.types import FieldAttr  # noqa
+from lamindb.core._data import add_labels
+from lamindb.core._feature_manager import parse_staged_feature_sets_from_anndata
+from lamindb.core._settings import settings
 from lamindb.models import (
     Artifact,
+    CanCurate,
+    Collection,
     Feature,
     Record,
     Run,
@@ -23,15 +54,25 @@ from lamindb.models import (
     ULabel,
 )
 
+from .._artifact import data_is_anndata
 from .._from_values import _format_values
-from ..core.exceptions import ValidationError
+from ..errors import InvalidArgument, ValidationError
 
 if TYPE_CHECKING:
-    from collections.abc import Iterable
+    from collections.abc import Iterable, MutableMapping
     from typing import Any
 
     from lamindb_setup.core.types import UPathStr
     from mudata import MuData
+    from spatialdata import SpatialData
+
+    from lamindb._query_set import RecordList
+
+
+def strip_ansi_codes(text):
+    # This pattern matches ANSI escape sequences
+    ansi_pattern = re.compile(r"\x1b\[[0-9;]*m")
+    return ansi_pattern.sub("", text)
 
 
 class CurateLookup:
@@ -40,8 +81,6 @@ class CurateLookup:
     Args:
         categoricals: A dictionary of categorical fields to lookup.
         slots: A dictionary of slot fields to lookup.
-        using_key: The key of the instance to lookup from. Defaults to the
-            current instance if not specified.
         public: Whether to lookup from the public instance. Defaults to False.
 
     Example:
@@ -55,48 +94,43 @@ class CurateLookup:
         self,
         categoricals: dict[str, FieldAttr],
         slots: dict[str, FieldAttr] = None,
-        using_key: str | None = None,
         public: bool = False,
     ) -> None:
         slots = slots or {}
-        self._fields = {**categoricals, **slots}
-        self._using_key = None if using_key == "default" else using_key
-        self._using_key_name = self._using_key or ln_setup.settings.instance.slug
+        self._categoricals = {**categoricals, **slots}
         self._public = public
-        debug_message = f"Lookup objects from {colors.italic(self._using_key_name)}"
-        logger.debug(debug_message)
 
     def __getattr__(self, name):
-        if name in self._fields:
-            registry = self._fields[name].field.model
+        if name in self._categoricals:
+            registry = self._categoricals[name].field.model
             if self._public and hasattr(registry, "public"):
                 return registry.public().lookup()
             else:
-                return get_registry_instance(registry, self._using_key).lookup()
+                return registry.lookup()
         raise AttributeError(
             f'"{self.__class__.__name__}" object has no attribute "{name}"'
         )
 
     def __getitem__(self, name):
-        if name in self._fields:
-            registry = self._fields[name].field.model
+        if name in self._categoricals:
+            registry = self._categoricals[name].field.model
             if self._public and hasattr(registry, "public"):
                 return registry.public().lookup()
             else:
-                return get_registry_instance(registry, self._using_key).lookup()
+                return registry.lookup()
         raise AttributeError(
             f'"{self.__class__.__name__}" object has no attribute "{name}"'
         )
 
     def __repr__(self) -> str:
-        if len(self._fields) > 0:
+        if len(self._categoricals) > 0:
             getattr_keys = "\n ".join(
-                [f".{key}" for key in self._fields if key.isidentifier()]
+                [f".{key}" for key in self._categoricals if key.isidentifier()]
             )
             getitem_keys = "\n ".join(
-                [str([key]) for key in self._fields if not key.isidentifier()]
+                [str([key]) for key in self._categoricals if not key.isidentifier()]
             )
-            ref = "public" if self._public else self._using_key_name
+            ref = "public" if self._public else "registries"
             return (
                 f"Lookup objects from the {colors.italic(ref)}:\n "
                 f"{colors.green(getattr_keys)}\n "
@@ -105,21 +139,422 @@ class CurateLookup:
                 "    → categories.alveolar_type_1_fibroblast_cell\n\n"
                 "To look up public ontologies, use .lookup(public=True)"
             )
-        else:  # pragma: no cover
+        else:  # pdagma: no cover
             return colors.warning("No fields are found!")
 
 
-class BaseCurator:
-    """Curate a dataset."""
+CAT_MANAGER_DOCSTRING = """Manage categoricals by updating registries."""
+
+
+VALIDATE_DOCSTRING = """Validate dataset.
+
+Raises:
+    lamindb.errors.ValidationError: If validation fails.
+"""
+
+SAVE_ARTIFACT_DOCSTRING = """Save an annotated artifact.
+
+Args:
+    key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`. Artifacts with the same key form a version family.
+    description: A description.
+    revises: Previous version of the artifact. Is an alternative way to passing `key` to trigger a new version.
+    run: The run that creates the artifact.
+
+Returns:
+    A saved artifact record.
+"""
+
+
+class Curator:
+    """Dataset curator.
+
+    A `Curator` object makes it easy to validate, standardize & annotate datasets.
+
+    See:
+        - :class:`~lamindb.curators.DataFrameCurator`
+        - :class:`~lamindb.curators.AnnDataCurator`
+    """
+
+    def __init__(self, dataset: Any, schema: Schema | None = None):
+        self._artifact: Artifact = None  # pass the dataset as an artifact
+        self._dataset: Any = dataset  # pass the dataset as a UPathStr or data object
+        if isinstance(self._dataset, Artifact):
+            self._artifact = self._dataset
+            if self._artifact.otype in {"DataFrame", "AnnData"}:
+                self._dataset = self._dataset.load()
+        self._schema: Schema | None = schema
+        self._is_validated: bool = False
+        self._cat_manager: CatManager = None  # is None for CatManager curators
+
+    @doc_args(VALIDATE_DOCSTRING)
+    def validate(self) -> bool | str:
+        """{}"""  # noqa: D415
+        pass  # pdagma: no cover
+
+    @doc_args(SAVE_ARTIFACT_DOCSTRING)
+    def save_artifact(
+        self,
+        *,
+        key: str | None = None,
+        description: str | None = None,
+        revises: Artifact | None = None,
+        run: Run | None = None,
+    ) -> Artifact:
+        """{}"""  # noqa: D415
+        # Note that this docstring has to be consistent with the Artifact()
+        # constructor signature
+        pass
+
+
+class DataFrameCurator(Curator):
+    # the example in the docstring is tested in test_curators_quickstart_example
+    """Curator for a DataFrame object.
+
+    See also :class:`~lamindb.Curator` and :class:`~lamindb.Schema`.
+
+    Args:
+        dataset: The DataFrame-like object to validate & annotate.
+        schema: A `Schema` object that defines the validation constraints.
+
+    Example::
+
+        import lamindb as ln
+        import bionty as bt
+
+        # define valid labels
+        cell_medium = ln.ULabel(name="CellMedium", is_type=True).save()
+        ln.ULabel(name="DMSO", type=cell_medium).save()
+        ln.ULabel(name="IFNG", type=cell_medium).save()
+        bt.CellType.from_source(name="B cell").save()
+        bt.CellType.from_source(name="T cell").save()
+
+        # define schema
+        schema = ln.Schema(
+            name="small_dataset1_obs_level_metadata",
+            features=[
+                ln.Feature(name="cell_medium", dtype="cat[ULabel[CellMedium]]").save(),
+                ln.Feature(name="sample_note", dtype=str).save(),
+                ln.Feature(name="cell_type_by_expert", dtype=bt.CellType).save(),
+                ln.Feature(name="cell_type_by_model", dtype=bt.CellType).save(),
+            ],
+        ).save()
+
+        # curate a DataFrame
+        df = datasets.small_dataset1(otype="DataFrame")
+        curator = ln.curators.DataFrameCurator(df, schema)
+        artifact = curator.save_artifact(key="example_datasets/dataset1.parquet")
+        assert artifact.schema == schema
+    """
+
+    def __init__(
+        self,
+        dataset: pd.DataFrame | Artifact,
+        schema: Schema,
+    ) -> None:
+        super().__init__(dataset=dataset, schema=schema)
+        if schema.n > 0:
+            # populate features
+            pandera_columns = {}
+            categoricals = {}
+            for feature in schema.features.all():
+                pandera_dtype = (
+                    feature.dtype if not feature.dtype.startswith("cat") else "category"
+                )
+                pandera_columns[feature.name] = pandera.Column(
+                    pandera_dtype, nullable=feature.nullable
+                )
+                if feature.dtype.startswith("cat"):
+                    categoricals[feature.name] = parse_dtype(feature.dtype)[0]["field"]
+            self._pandera_schema = pandera.DataFrameSchema(
+                pandera_columns, coerce=schema.coerce_dtype
+            )
+            # now deal with detailed validation of categoricals
+            self._cat_manager = DataFrameCatManager(
+                self._dataset,
+                categoricals=categoricals,
+            )
+        else:
+            assert schema.itype is not None  # noqa: S101
+
+    @property
+    @doc_args(CAT_MANAGER_DOCSTRING)
+    def cat(self) -> CatManager:
+        """{}"""  # noqa: D415
+        return self._cat_manager
+
+    def standardize(self) -> None:
+        """Standardize the dataset.
+
+        - Adds missing columns if a default value for a feature is defined.
+        - Fills missing values with the default value if a default value for a feature is defined.
+        """
+        for feature in self._schema.members:
+            if feature.name not in self._dataset.columns:
+                if feature.default_value is not None:
+                    self._dataset[feature.name] = feature.default_value
+                else:
+                    raise ValidationError(
+                        f"Missing column {feature.name} cannot be added because no default value is defined for this feature"
+                    )
+            else:
+                if feature.default_value is not None:
+                    if isinstance(
+                        self._dataset[feature.name].dtype, pd.CategoricalDtype
+                    ):
+                        if (
+                            feature.default_value
+                            not in self._dataset[feature.name].cat.categories
+                        ):
+                            self._dataset[feature.name] = self._dataset[
+                                feature.name
+                            ].cat.add_categories(feature.default_value)
+                    self._dataset[feature.name] = self._dataset[feature.name].fillna(
+                        feature.default_value
+                    )
+
+    @doc_args(VALIDATE_DOCSTRING)
+    def validate(self) -> None:
+        """{}"""  # noqa: D415
+        if self._schema.n > 0:
+            self._cat_manager.validate()
+            try:
+                self._pandera_schema.validate(self._dataset)
+                if self._cat_manager._is_validated:
+                    self._is_validated = True
+                else:
+                    self._is_validated = False
+                    raise ValidationError(
+                        self._cat_manager._validate_category_error_messages
+                    )
+            except pandera.errors.SchemaError as err:
+                self._is_validated = False
+                # .exconly() doesn't exist on SchemaError
+                raise ValidationError(str(err)) from err
+        else:
+            result = parse_dtype_single_cat(self._schema.itype, is_itype=True)
+            registry: CanCurate = result["registry"]
+            inspector = registry.inspect(
+                self._dataset.columns,
+                result["field"],
+                mute=True,
+            )
+            if len(inspector.non_validated) > 0:
+                # also check public ontology
+                if hasattr(registry, "public"):
+                    registry.from_values(
+                        inspector.non_validated, result["field"], mute=True
+                    ).save()
+                    inspector = registry.inspect(
+                        inspector.non_validated, result["field"], mute=True
+                    )
+                if len(inspector.non_validated) > 0:
+                    self._is_validated = False
+                    raise ValidationError(
+                        f"Invalid identifiers for {self._schema.itype}: {inspector.non_validated}"
+                    )
+
+    @doc_args(SAVE_ARTIFACT_DOCSTRING)
+    def save_artifact(
+        self,
+        *,
+        key: str | None = None,
+        description: str | None = None,
+        revises: Artifact | None = None,
+        run: Run | None = None,
+    ):
+        """{}"""  # noqa: D415
+        if not self._is_validated:
+            self.validate()  # raises ValidationError if doesn't validate
+        result = parse_dtype_single_cat(self._schema.itype, is_itype=True)
+        return save_artifact(  # type: ignore
+            self._dataset,
+            description=description,
+            fields=self._cat_manager.categoricals,
+            columns_field=result["field"],
+            key=key,
+            artifact=self._artifact,
+            revises=revises,
+            run=run,
+            schema=self._schema,
+        )
+
+
+class AnnDataCurator(Curator):
+    # the example in the docstring is tested in test_curators_quickstart_example
+    """Curator for a DataFrame object.
+
+    See also :class:`~lamindb.Curator` and :class:`~lamindb.Schema`.
+
+    Args:
+        dataset: The AnnData-like object to validate & annotate.
+        schema: A `Schema` object that defines the validation constraints.
+
+    Example::
+
+        import lamindb as ln
+        import bionty as bt
+
+        # define valid labels
+        cell_medium = ln.ULabel(name="CellMedium", is_type=True).save()
+        ln.ULabel(name="DMSO", type=cell_medium).save()
+        ln.ULabel(name="IFNG", type=cell_medium).save()
+        bt.CellType.from_source(name="B cell").save()
+        bt.CellType.from_source(name="T cell").save()
+
+        # define obs schema
+        obs_schema = ln.Schema(
+            name="small_dataset1_obs_level_metadata",
+            features=[
+                ln.Feature(name="cell_medium", dtype="cat[ULabel[CellMedium]]").save(),
+                ln.Feature(name="sample_note", dtype=str).save(),
+                ln.Feature(name="cell_type_by_expert", dtype=bt.CellType").save(),
+                ln.Feature(name="cell_type_by_model", dtype=bt.CellType").save(),
+            ],
+        ).save()
+
+        # define var schema
+        var_schema = ln.Schema(
+            name="scRNA_seq_var_schema",
+            itype=bt.Gene.ensembl_gene_id,
+            dtype="num",
+        ).save()
+
+        # define composite schema
+        anndata_schema = ln.Schema(
+            name="small_dataset1_anndata_schema",
+            otype="AnnData",
+            components={"obs": obs_schema, "var": var_schema},
+        ).save()
+
+        # curate an AnnData
+        adata = datasets.small_dataset1(otype="AnnData")
+        curator = ln.curators.AnnDataCurator(adata, anndata_schema)
+        artifact = curator.save_artifact(key="example_datasets/dataset1.h5ad")
+        assert artifact.schema == anndata_schema
+    """
+
+    def __init__(
+        self,
+        dataset: AnnData | Artifact,
+        schema: Schema,
+    ) -> None:
+        super().__init__(dataset=dataset, schema=schema)
+        if not data_is_anndata(self._dataset):
+            raise InvalidArgument("dataset must be AnnData-like.")
+        if schema.otype != "AnnData":
+            raise InvalidArgument("Schema otype must be 'AnnData'.")
+        self._obs_curator = DataFrameCurator(
+            self._dataset.obs, schema._get_component("obs")
+        )
+        self._var_curator = DataFrameCurator(
+            self._dataset.var.T, schema._get_component("var")
+        )
+
+    @doc_args(VALIDATE_DOCSTRING)
+    def validate(self) -> None:
+        """{}"""  # noqa: D415
+        self._obs_curator.validate()
+        self._var_curator.validate()
+        self._is_validated = True
+
+    @doc_args(SAVE_ARTIFACT_DOCSTRING)
+    def save_artifact(self, *, key=None, description=None, revises=None, run=None):
+        """{}"""  # noqa: D415
+        if not self._is_validated:
+            self.validate()  # raises ValidationError if doesn't validate
+        result = parse_dtype_single_cat(self._var_curator._schema.itype, is_itype=True)
+        return save_artifact(  # type: ignore
+            self._dataset,
+            description=description,
+            fields=self._obs_curator._cat_manager.categoricals,
+            columns_field=result["field"],
+            key=key,
+            artifact=self._artifact,
+            revises=revises,
+            run=run,
+            schema=self._schema,
+        )
+
+
+class CatManager:
+    """Manage valid categoricals by updating registries.
+
+    A `CatManager` object makes it easy to validate, standardize & annotate datasets.
+
+    Example:
+
+    >>> cat_manager = ln.CatManager(
+    >>>     dataset,
+    >>>     # define validation criteria as mappings
+    >>>     columns=Feature.name,  # map column names
+    >>>     categoricals={"perturbation": ULabel.name},  # map categories
+    >>> )
+    >>> cat_manager.validate()  # validate the dataframe
+    >>> artifact = cat_manager.save_artifact(description="my RNA-seq")
+    >>> artifact.describe()  # see annotations
+
+    `cat_manager.validate()` maps values within `df` according to the mapping criteria and logs validated & problematic values.
+
+    If you find non-validated values, you have several options:
+
+    - new values found in the data can be registered using :meth:`~lamindb.core.DataFrameCatManager.add_new_from`
+    - non-validated values can be accessed using :meth:`~lamindb.core.DataFrameCatManager.non_validated` and addressed manually
+    """
+
+    def __init__(
+        self, *, dataset, categoricals, sources, organism, exclude, columns_field=None
+    ):
+        # the below is shared with Curator
+        self._artifact: Artifact = None  # pass the dataset as an artifact
+        self._dataset: Any = dataset  # pass the dataset as a UPathStr or data object
+        if isinstance(self._dataset, Artifact):
+            self._artifact = self._dataset
+            if self._artifact.otype in {"DataFrame", "AnnData"}:
+                self._dataset = self._dataset.load()
+        self._is_validated: bool = False
+        # shared until here
+        self._categoricals = categoricals or {}
+        self._non_validated = None
+        self._organism = organism
+        self._sources = sources or {}
+        self._exclude = exclude or {}
+        self._columns_field = columns_field
+        self._validate_category_error_messages: str = ""
+
+    @property
+    def non_validated(self) -> dict[str, list[str]]:
+        """Return the non-validated features and labels."""
+        if self._non_validated is None:
+            raise ValidationError("Please run validate() first!")
+        return self._non_validated
 
-    def __init_subclass__(cls, **kwargs):
-        super().__init_subclass__(**kwargs)
-        import sys
+    @property
+    def categoricals(self) -> dict:
+        """Return the columns fields to validate against."""
+        return self._categoricals
 
-        # Deprecated methods
-        if "sphinx" not in sys.modules:
-            if hasattr(cls, "_add_new_from_columns"):
-                cls.add_new_from_columns = cls._add_new_from_columns
+    def _replace_synonyms(
+        self, key: str, syn_mapper: dict, values: pd.Series | pd.Index
+    ):
+        # replace the values in df
+        std_values = values.map(lambda unstd_val: syn_mapper.get(unstd_val, unstd_val))
+        # remove the standardized values from self.non_validated
+        non_validated = [i for i in self.non_validated[key] if i not in syn_mapper]
+        if len(non_validated) == 0:
+            self._non_validated.pop(key, None)  # type: ignore
+        else:
+            self._non_validated[key] = non_validated  # type: ignore
+        # logging
+        n = len(syn_mapper)
+        if n > 0:
+            syn_mapper_print = _format_values(
+                [f'"{k}" → "{v}"' for k, v in syn_mapper.items()], sep=""
+            )
+            s = "s" if n > 1 else ""
+            logger.success(
+                f'standardized {n} synonym{s} in "{key}": {colors.green(syn_mapper_print)}'
+            )
+        return std_values
 
     def validate(self) -> bool:
         """Validate dataset.
@@ -127,9 +562,9 @@ class BaseCurator:
         This method also registers the validated records in the current instance.
 
         Returns:
-            Boolean indicating whether the dataset is validated.
+            The boolean `True` if the dataset is validated. Otherwise, a string with the error message.
         """
-        pass  # pragma: no cover
+        pass
 
     def standardize(self, key: str) -> None:
         """Replace synonyms with standardized values.
@@ -142,30 +577,48 @@ class BaseCurator:
         Returns:
             None
         """
-        pass  # pragma: no cover
+        pass  # pdagma: no cover
 
+    @doc_args(SAVE_ARTIFACT_DOCSTRING)
     def save_artifact(
         self,
-        description: str | None = None,
+        *,
         key: str | None = None,
+        description: str | None = None,
         revises: Artifact | None = None,
         run: Run | None = None,
     ) -> Artifact:
-        """Save the dataset as artifact.
+        """{}"""  # noqa: D415
+        from lamindb.core._settings import settings
 
-        Args:
-            description: A description of the DataFrame object.
-            key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`. Artifacts with the same key form a revision family.
-            revises: Previous version of the artifact. Triggers a revision.
-            run: The run that creates the artifact.
+        if not self._is_validated:
+            self.validate()  # returns True or False
+            if not self._is_validated:  # need to raise error manually
+                raise ValidationError("Dataset does not validate. Please curate.")
 
-        Returns:
-            A saved artifact record.
-        """
-        pass  # pragma: no cover
+        # Make sure all labels are saved in the current instance
+        verbosity = settings.verbosity
+        try:
+            settings.verbosity = "warning"
+            self._artifact = save_artifact(  # type: ignore
+                self._dataset,
+                description=description,
+                fields=self.categoricals,
+                columns_field=self._columns_field,
+                key=key,
+                artifact=self._artifact,
+                revises=revises,
+                run=run,
+                schema=None,
+                organism=self._organism,
+            )
+        finally:
+            settings.verbosity = verbosity
+
+        return self._artifact
 
 
-class DataFrameCurator(BaseCurator):
+class DataFrameCatManager(CatManager):
     """Curation flow for a DataFrame object.
 
     See also :class:`~lamindb.Curator`.
@@ -174,7 +627,6 @@ class DataFrameCurator(BaseCurator):
         df: The DataFrame object to curate.
         columns: The field attribute for the feature column.
         categoricals: A dictionary mapping column names to registry_field.
-        using_key: The reference instance containing registries to validate against.
         verbosity: The verbosity level.
         organism: The organism name.
         sources: A dictionary mapping column names to Source records.
@@ -191,165 +643,103 @@ class DataFrameCurator(BaseCurator):
         ...     df,
         ...     categoricals={
         ...         "cell_type_ontology_id": bt.CellType.ontology_id,
-        ...         "donor_id": ln.ULabel.name
+        ...         "donor_id": ULabel.name
         ...     }
         ... )
     """
 
     def __init__(
         self,
-        df: pd.DataFrame,
+        df: pd.DataFrame | Artifact,
         columns: FieldAttr = Feature.name,
         categoricals: dict[str, FieldAttr] | None = None,
-        using_key: str | None = None,
         verbosity: str = "hint",
         organism: str | None = None,
         sources: dict[str, Record] | None = None,
         exclude: dict | None = None,
-        check_valid_keys: bool = True,
     ) -> None:
         from lamindb.core._settings import settings
 
         if organism is not None and not isinstance(organism, str):
             raise ValueError("organism must be a string such as 'human' or 'mouse'!")
 
-        self._df = df
-        self._fields = categoricals or {}
-        self._columns_field = columns
-        self._using_key = using_key
-        # TODO: change verbosity back
         settings.verbosity = verbosity
-        self._artifact = None
-        self._collection = None
-        self._validated = False
-        self._kwargs = {"organism": organism} if organism else {}
-        self._sources = sources or {}
-        self._exclude = exclude or {}
         self._non_validated = None
-        if check_valid_keys:
-            self._check_valid_keys()
+        super().__init__(
+            dataset=df,
+            columns_field=columns,
+            organism=organism,
+            categoricals=categoricals,
+            sources=sources,
+            exclude=exclude,
+        )
         self._save_columns()
 
-    @property
-    def non_validated(self) -> dict[str, list[str]]:
-        """Return the non-validated features and labels."""
-        if self._non_validated is None:
-            raise ValidationError("Please run validate() first!")
-        return self._non_validated
-
-    @property
-    def fields(self) -> dict:
-        """Return the columns fields to validate against."""
-        return self._fields
-
-    def lookup(
-        self, using_key: str | None = None, public: bool = False
-    ) -> CurateLookup:
+    def lookup(self, public: bool = False) -> CurateLookup:
         """Lookup categories.
 
         Args:
-            using_key: The instance where the lookup is performed.
-                if "public", the lookup is performed on the public reference.
+            public: If "public", the lookup is performed on the public reference.
         """
         return CurateLookup(
-            categoricals=self._fields,
+            categoricals=self._categoricals,
             slots={"columns": self._columns_field},
-            using_key=using_key or self._using_key,
             public=public,
         )
 
-    def _check_valid_keys(self, extra: set | None = None) -> None:
-        extra = extra or set()
-        for name, d in {
-            "categoricals": self._fields,
-            "sources": self._sources,
-            "exclude": self._exclude,
-        }.items():
-            if not isinstance(d, dict):
-                raise TypeError(f"{name} must be a dictionary!")
-            valid_keys = set(self._df.columns) | {"columns"} | extra
-            nonval_keys = [key for key in d.keys() if key not in valid_keys]
-            n = len(nonval_keys)
-            s = "s" if n > 1 else ""
-            are = "are" if n > 1 else "is"
-            if len(nonval_keys) > 0:
-                raise ValidationError(
-                    f"key{s} passed to {name} {are} not present in columns: {colors.yellow(_format_values(nonval_keys))}"
-                )
-
     def _save_columns(self, validated_only: bool = True) -> None:
         """Save column name records."""
         # Always save features specified as the fields keys
         update_registry(
-            values=list(self.fields.keys()),
+            values=list(self.categoricals.keys()),
             field=self._columns_field,
             key="columns",
-            using_key=self._using_key,
             validated_only=False,
             source=self._sources.get("columns"),
             exclude=self._exclude.get("columns"),
-            **self._kwargs,  # type: ignore
         )
 
         # Save the rest of the columns based on validated_only
-        additional_columns = set(self._df.columns) - set(self.fields.keys())
+        additional_columns = set(self._dataset.columns) - set(self.categoricals.keys())
         if additional_columns:
             update_registry(
                 values=list(additional_columns),
                 field=self._columns_field,
                 key="columns",
-                using_key=self._using_key,
                 validated_only=validated_only,
-                df=self._df,  # Get the Feature type from df
+                df=self._dataset,  # Get the Feature type from df
                 source=self._sources.get("columns"),
                 exclude=self._exclude.get("columns"),
-                **self._kwargs,  # type: ignore
             )
 
-    def add_new_from(self, key: str, organism: str | None = None, **kwargs):
-        """Add validated & new categories.
+    @deprecated(new_name="is run by default")
+    def add_new_from_columns(self, organism: str | None = None, **kwargs):
+        pass
+
+    def validate(self) -> bool:
+        """Validate variables and categorical observations.
+
+        This method also registers the validated records in the current instance:
+        - from public sources
 
         Args:
-            key: The key referencing the slot in the DataFrame from which to draw terms.
             organism: The organism name.
-            **kwargs: Additional keyword arguments to pass to create new records
+
+        Returns:
+            Whether the DataFrame is validated.
         """
-        if len(kwargs) > 0 and key == "all":
-            raise ValueError("Cannot pass additional arguments to 'all' key!")
-        self._kwargs.update({"organism": organism} if organism else {})
-        self._update_registry(key, validated_only=False, **self._kwargs, **kwargs)
-
-    def _add_new_from_columns(self, organism: str | None = None, **kwargs):
-        """Deprecated to run by default during init."""
-        warnings.warn(
-            "`.add_new_from_columns()` is deprecated and will be removed in a future version. It's run by default during initialization.",
-            DeprecationWarning,
-            stacklevel=2,
+        # add all validated records to the current instance
+        self._update_registry_all()
+        self._validate_category_error_messages = ""  # reset the error messages
+        self._is_validated, self._non_validated = validate_categories_in_df(  # type: ignore
+            self._dataset,
+            fields=self.categoricals,
+            sources=self._sources,
+            exclude=self._exclude,
+            curator=self,
+            organism=self._organism,
         )
-        pass
-
-    def _replace_synonyms(
-        self, key: str, syn_mapper: dict, values: pd.Series | pd.Index
-    ):
-        # replace the values in df
-        std_values = values.map(lambda unstd_val: syn_mapper.get(unstd_val, unstd_val))
-        # remove the standardized values from self.non_validated
-        non_validated = [i for i in self.non_validated[key] if i not in syn_mapper]
-        if len(non_validated) == 0:
-            self._non_validated.pop(key, None)  # type: ignore
-        else:
-            self._non_validated[key] = non_validated  # type: ignore
-        # logging
-        n = len(syn_mapper)
-        if n > 0:
-            syn_mapper_print = _format_values(
-                [f'"{k}" → "{v}"' for k, v in syn_mapper.items()], sep=""
-            )
-            s = "s" if n > 1 else ""
-            logger.success(
-                f'standardized {n} synonym{s} in "{key}": {colors.green(syn_mapper_print)}'
-            )
-        return std_values
+        return self._is_validated
 
     def standardize(self, key: str) -> None:
         """Replace synonyms with standardized values.
@@ -359,6 +749,8 @@ class DataFrameCurator(BaseCurator):
         Args:
             key: The key referencing the column in the DataFrame to standardize.
         """
+        if self._artifact is not None:
+            raise RuntimeError("can't mutate the dataset when an artifact is passed!")
         # list is needed to avoid RuntimeError: dictionary changed size during iteration
         avail_keys = list(self.non_validated.keys())
         if len(avail_keys) == 0:
@@ -367,137 +759,74 @@ class DataFrameCurator(BaseCurator):
 
         if key == "all":
             for k in avail_keys:
-                if k in self._fields:  # needed to exclude var_index
+                if k in self._categoricals:  # needed to exclude var_index
                     syn_mapper = standardize_categories(
                         self.non_validated[k],
-                        field=self._fields[k],
-                        using_key=self._using_key,
+                        field=self._categoricals[k],
                         source=self._sources.get(k),
-                        **self._kwargs,
                     )
-                    self._df[k] = self._replace_synonyms(k, syn_mapper, self._df[k])
+                    self._dataset[k] = self._replace_synonyms(
+                        k, syn_mapper, self._dataset[k]
+                    )
         else:
             if key not in avail_keys:
-                if key in self._fields:
+                if key in self._categoricals:
                     logger.info(f"No unstandardized values found for {key!r}")
                 else:
                     raise KeyError(
                         f"{key!r} is not a valid key, available keys are: {_format_values(avail_keys)}!"
                     )
             else:
-                if key in self._fields:  # needed to exclude var_index
+                if key in self._categoricals:  # needed to exclude var_index
                     syn_mapper = standardize_categories(
                         self.non_validated[key],
-                        field=self._fields[key],
-                        using_key=self._using_key,
+                        field=self._categoricals[key],
                         source=self._sources.get(key),
-                        **self._kwargs,
+                        organism=self._organism,
                     )
-                    self._df[key] = self._replace_synonyms(
-                        key, syn_mapper, self._df[key]
+                    self._dataset[key] = self._replace_synonyms(
+                        key, syn_mapper, self._dataset[key]
                     )
 
+    def _update_registry_all(self, validated_only: bool = True, **kwargs):
+        """Save labels for all features."""
+        for name in self.categoricals.keys():
+            self._update_registry(name, validated_only=validated_only, **kwargs)
+
     def _update_registry(
         self, categorical: str, validated_only: bool = True, **kwargs
     ) -> None:
         if categorical == "all":
             self._update_registry_all(validated_only=validated_only, **kwargs)
         else:
-            if categorical not in self.fields:
+            if categorical not in self.categoricals:
                 raise ValidationError(
                     f"Feature {categorical} is not part of the fields!"
                 )
             update_registry(
-                values=_flatten_unique(self._df[categorical]),
-                field=self.fields[categorical],
+                values=_flatten_unique(self._dataset[categorical]),
+                field=self.categoricals[categorical],
                 key=categorical,
-                using_key=self._using_key,
                 validated_only=validated_only,
                 source=self._sources.get(categorical),
                 exclude=self._exclude.get(categorical),
-                **kwargs,
+                organism=self._organism,
             )
             # adding new records removes them from non_validated
             if not validated_only and self._non_validated:
                 self._non_validated.pop(categorical, None)  # type: ignore
 
-    def _update_registry_all(self, validated_only: bool = True, **kwargs):
-        """Save labels for all features."""
-        for name in self.fields.keys():
-            self._update_registry(name, validated_only=validated_only, **kwargs)
-
-    def validate(self, organism: str | None = None) -> bool:
-        """Validate variables and categorical observations.
-
-        This method also registers the validated records in the current instance:
-        - from public sources
-        - from the using_key instance
+    def add_new_from(self, key: str, **kwargs):
+        """Add validated & new categories.
 
         Args:
+            key: The key referencing the slot in the DataFrame from which to draw terms.
             organism: The organism name.
-
-        Returns:
-            Whether the DataFrame is validated.
-        """
-        self._kwargs.update({"organism": organism} if organism else {})
-
-        # add all validated records to the current instance
-        self._update_registry_all()
-
-        self._validated, self._non_validated = validate_categories_in_df(  # type: ignore
-            self._df,
-            fields=self.fields,
-            using_key=self._using_key,
-            sources=self._sources,
-            exclude=self._exclude,
-            **self._kwargs,
-        )
-        return self._validated
-
-    def save_artifact(
-        self,
-        description: str | None = None,
-        key: str | None = None,
-        revises: Artifact | None = None,
-        run: Run | None = None,
-    ) -> Artifact:
-        """Save the validated DataFrame and metadata.
-
-        Args:
-            description: Description of the DataFrame object.
-            key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`.
-                Artifacts with the same key form a revision family.
-            revises: Previous version of the artifact. Triggers a revision.
-            run: The run that creates the artifact.
-
-        Returns:
-            A saved artifact record.
+            **kwargs: Additional keyword arguments to pass to create new records
         """
-        from lamindb.core._settings import settings
-
-        if not self._validated:
-            self.validate()
-            if not self._validated:
-                raise ValidationError("Dataset does not validate. Please curate.")
-
-        # Make sure all labels are saved in the current instance
-        verbosity = settings.verbosity
-        try:
-            settings.verbosity = "warning"
-            self._artifact = save_artifact(
-                self._df,
-                description=description,
-                fields=self.fields,
-                columns_field=self._columns_field,
-                key=key,
-                revises=revises,
-                run=run,
-                **self._kwargs,
-            )
-        finally:
-            settings.verbosity = verbosity
-
-        return self._artifact
+        if len(kwargs) > 0 and key == "all":
+            raise ValueError("Cannot pass additional arguments to 'all' key!")
+        self._update_registry(key, validated_only=False, **kwargs)
 
     def clean_up_failed_runs(self):
         """Clean up previous failed runs that don't save any outputs."""
@@ -509,21 +838,14 @@ class DataFrameCurator(BaseCurator):
             ).delete()
 
 
-class AnnDataCurator(DataFrameCurator):
-    """Curation flow for ``AnnData``.
-
-    See also :class:`~lamindb.Curator`.
-
-    Note that if genes are removed from the AnnData object, the object should be recreated using :meth:`~lamindb.Curator.from_anndata`.
-
-    See :doc:`docs:cellxgene-curate` for instructions on how to curate against a specific cellxgene schema version.
+class AnnDataCatManager(CatManager):
+    """Manage categorical curation.
 
     Args:
         data: The AnnData object or an AnnData-like path.
         var_index: The registry field for mapping the ``.var`` index.
         categoricals: A dictionary mapping ``.obs.columns`` to a registry field.
         obs_columns: The registry field for mapping the ``.obs.columns``.
-        using_key: A reference LaminDB instance.
         verbosity: The verbosity level.
         organism: The organism name.
         sources: A dictionary mapping ``.obs.columns`` to Source records.
@@ -538,7 +860,7 @@ class AnnDataCurator(DataFrameCurator):
         ...     var_index=bt.Gene.ensembl_gene_id,
         ...     categoricals={
         ...         "cell_type_ontology_id": bt.CellType.ontology_id,
-        ...         "donor_id": ln.ULabel.name
+        ...         "donor_id": ULabel.name
         ...     },
         ...     organism="human",
         ... )
@@ -546,56 +868,48 @@ class AnnDataCurator(DataFrameCurator):
 
     def __init__(
         self,
-        data: ad.AnnData | UPathStr,
+        data: ad.AnnData | Artifact,
         var_index: FieldAttr,
         categoricals: dict[str, FieldAttr] | None = None,
         obs_columns: FieldAttr = Feature.name,
-        using_key: str | None = None,
         verbosity: str = "hint",
         organism: str | None = None,
         sources: dict[str, Record] | None = None,
         exclude: dict | None = None,
     ) -> None:
-        from lamindb_setup.core import upath
-
         if isinstance(var_index, str):
             raise TypeError("var_index parameter has to be a bionty field")
 
-        from .._artifact import data_is_anndata
-
         if sources is None:
             sources = {}
         if not data_is_anndata(data):
-            raise TypeError(
-                "data has to be an AnnData object or a path to AnnData-like"
-            )
-        if isinstance(data, ad.AnnData):
-            self._adata = data
-        else:  # pragma: no cover
-            from lamindb.core.storage._backed_access import backed_access
-
-            self._adata = backed_access(upath.create_path(data))
+            raise TypeError("data has to be an AnnData object")
 
         if "symbol" in str(var_index):
             logger.warning(
                 "indexing datasets with gene symbols can be problematic: https://docs.lamin.ai/faq/symbol-mapping"
             )
 
-        self._data = data
+        self._obs_fields = categoricals or {}
         self._var_field = var_index
         super().__init__(
-            df=self._adata.obs,
+            dataset=data,
             categoricals=categoricals,
+            sources=sources,
+            organism=organism,
+            exclude=exclude,
+            columns_field=var_index,
+        )
+        self._adata = self._dataset
+        self._obs_df_curator = DataFrameCatManager(
+            df=self._adata.obs,
+            categoricals=self.categoricals,
             columns=obs_columns,
-            using_key=using_key,
             verbosity=verbosity,
-            organism=organism,
+            organism=None,
             sources=sources,
             exclude=exclude,
-            check_valid_keys=False,
         )
-        self._obs_fields = categoricals or {}
-        self._check_valid_keys(extra={"var_index"})
 
     @property
     def var_index(self) -> FieldAttr:
@@ -607,54 +921,53 @@ class AnnDataCurator(DataFrameCurator):
         """Return the obs fields to validate against."""
         return self._obs_fields
 
-    def lookup(
-        self, using_key: str | None = None, public: bool = False
-    ) -> CurateLookup:
+    def lookup(self, public: bool = False) -> CurateLookup:
         """Lookup categories.
 
         Args:
-            using_key: The instance where the lookup is performed.
-                if "public", the lookup is performed on the public reference.
+            public: If "public", the lookup is performed on the public reference.
         """
         return CurateLookup(
             categoricals=self._obs_fields,
             slots={"columns": self._columns_field, "var_index": self._var_field},
-            using_key=using_key or self._using_key,
             public=public,
         )
 
     def _save_from_var_index(
-        self, validated_only: bool = True, organism: str | None = None
+        self,
+        validated_only: bool = True,
     ):
         """Save variable records."""
         update_registry(
             values=list(self._adata.var.index),
             field=self.var_index,
             key="var_index",
-            using_key=self._using_key,
             validated_only=validated_only,
-            organism=organism,
+            organism=self._organism,
             source=self._sources.get("var_index"),
             exclude=self._exclude.get("var_index"),
         )
 
-    def _update_registry_all(self, validated_only: bool = True, **kwargs):
-        """Save labels for all features."""
-        self._save_from_var_index(validated_only=validated_only, **self._kwargs)
-        for name in self._obs_fields.keys():
-            self._update_registry(name, validated_only=validated_only, **self._kwargs)
+    def add_new_from(self, key: str, **kwargs):
+        """Add validated & new categories.
 
-    def add_new_from_var_index(self, organism: str | None = None, **kwargs):
+        Args:
+            key: The key referencing the slot in the DataFrame from which to draw terms.
+            organism: The organism name.
+            **kwargs: Additional keyword arguments to pass to create new records
+        """
+        self._obs_df_curator.add_new_from(key, **kwargs)
+
+    def add_new_from_var_index(self, **kwargs):
         """Update variable records.
 
         Args:
             organism: The organism name.
             **kwargs: Additional keyword arguments to pass to create new records.
         """
-        self._kwargs.update({"organism": organism} if organism else {})
-        self._save_from_var_index(validated_only=False, **self._kwargs, **kwargs)
+        self._save_from_var_index(validated_only=False, **kwargs)
 
-    def validate(self, organism: str | None = None) -> bool:
+    def validate(self) -> bool:
         """Validate categories.
 
         This method also registers the validated records in the current instance.
@@ -665,38 +978,25 @@ class AnnDataCurator(DataFrameCurator):
         Returns:
             Whether the AnnData object is validated.
         """
-        self._kwargs.update({"organism": organism} if organism else {})
-        if self._using_key is not None and self._using_key != "default":
-            logger.important(
-                f"validating metadata using registries of instance {colors.italic(self._using_key)}"
-            )
+        self._validate_category_error_messages = ""  # reset the error messages
 
         # add all validated records to the current instance
-        self._update_registry_all()
-
+        self._save_from_var_index(validated_only=True)
         validated_var, non_validated_var = validate_categories(
             self._adata.var.index,
             field=self._var_field,
             key="var_index",
-            using_key=self._using_key,
             source=self._sources.get("var_index"),
             hint_print=".add_new_from_var_index()",
             exclude=self._exclude.get("var_index"),
-            **self._kwargs,  # type: ignore
-        )
-        validated_obs, non_validated_obs = validate_categories_in_df(
-            self._adata.obs,
-            fields=self.categoricals,
-            using_key=self._using_key,
-            sources=self._sources,
-            exclude=self._exclude,
-            **self._kwargs,
+            organism=self._organism,  # type: ignore
         )
-        self._non_validated = non_validated_obs  # type: ignore
+        validated_obs = self._obs_df_curator.validate()
+        self._non_validated = self._obs_df_curator._non_validated  # type: ignore
         if len(non_validated_var) > 0:
             self._non_validated["var_index"] = non_validated_var  # type: ignore
-        self._validated = validated_var and validated_obs
-        return self._validated
+        self._is_validated = validated_var and validated_obs
+        return self._is_validated
 
     def standardize(self, key: str):
         """Replace synonyms with standardized values.
@@ -709,68 +1009,26 @@ class AnnDataCurator(DataFrameCurator):
 
         Inplace modification of the dataset.
         """
+        if self._artifact is not None:
+            raise RuntimeError("can't mutate the dataset when an artifact is passed!")
         if key in self._adata.obs.columns or key == "all":
             # standardize obs columns
-            super().standardize(key)
+            self._obs_df_curator.standardize(key)
         # in addition to the obs columns, standardize the var.index
         if key == "var_index" or key == "all":
             syn_mapper = standardize_categories(
                 self._adata.var.index,
                 field=self.var_index,
-                using_key=self._using_key,
                 source=self._sources.get("var_index"),
-                **self._kwargs,
+                organism=self._organism,
             )
             if "var_index" in self._non_validated:  # type: ignore
                 self._adata.var.index = self._replace_synonyms(
                     "var_index", syn_mapper, self._adata.var.index
                 )
 
-    def save_artifact(
-        self,
-        description: str | None = None,
-        key: str | None = None,
-        revises: Artifact | None = None,
-        run: Run | None = None,
-    ) -> Artifact:
-        """Save the validated ``AnnData`` and metadata.
-
-        Args:
-            description: A description of the ``AnnData`` object.
-            key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`.
-                Artifacts with the same key form a revision family.
-            revises: Previous version of the artifact. Triggers a revision.
-            run: The run that creates the artifact.
-
-        Returns:
-            A saved artifact record.
-        """
-        from lamindb.core._settings import settings
-
-        if not self._validated:
-            self.validate()
-            if not self._validated:
-                raise ValidationError("Dataset does not validate. Please curate.")
-        verbosity = settings.verbosity
-        try:
-            settings.verbosity = "warning"
-            self._artifact = save_artifact(
-                self._data,
-                adata=self._adata,
-                description=description,
-                columns_field=self.var_index,
-                fields=self.categoricals,
-                key=key,
-                revises=revises,
-                run=run,
-                **self._kwargs,
-            )
-        finally:
-            settings.verbosity = verbosity
-        return self._artifact
-
 
-class MuDataCurator:
+class MuDataCatManager(CatManager):
     """Curation flow for a ``MuData`` object.
 
     See also :class:`~lamindb.Curator`.
@@ -782,10 +1040,9 @@ class MuDataCurator:
         mdata: The MuData object to curate.
         var_index: The registry field for mapping the ``.var`` index for each modality.
             For example:
-            ``{"modality_1": bt.Gene.ensembl_gene_id, "modality_2": ln.CellMarker.name}``
+            ``{"modality_1": bt.Gene.ensembl_gene_id, "modality_2": CellMarker.name}``
         categoricals: A dictionary mapping ``.obs.columns`` to a registry field.
             Use modality keys to specify categoricals for MuData slots such as `"rna:cell_type": bt.CellType.name"`.
-        using_key: A reference LaminDB instance.
         verbosity: The verbosity level.
         organism: The organism name.
         sources: A dictionary mapping ``.obs.columns`` to Source records.
@@ -799,11 +1056,11 @@ class MuDataCurator:
         ...     mdata,
         ...     var_index={
         ...         "rna": bt.Gene.ensembl_gene_id,
-        ...         "adt": ln.CellMarker.name
+        ...         "adt": CellMarker.name
         ...     },
         ...     categoricals={
         ...         "cell_type_ontology_id": bt.CellType.ontology_id,
-        ...         "donor_id": ln.ULabel.name
+        ...         "donor_id": ULabel.name
         ...     },
         ...     organism="human",
         ... )
@@ -811,52 +1068,47 @@ class MuDataCurator:
 
     def __init__(
         self,
-        mdata: MuData,
+        mdata: MuData | Artifact,
         var_index: dict[str, FieldAttr],
         categoricals: dict[str, FieldAttr] | None = None,
-        using_key: str | None = None,
         verbosity: str = "hint",
         organism: str | None = None,
         sources: dict[str, Record] | None = None,
         exclude: dict | None = None,  # {modality: {field: [values]}}
     ) -> None:
-        if sources is None:
-            sources = {}
-        self._sources = sources
-        if exclude is None:
-            exclude = {}
-        self._exclude = exclude
-        self._mdata = mdata
-        self._kwargs = {"organism": organism} if organism else {}
+        super().__init__(
+            dataset=mdata,
+            categoricals={},
+            sources=sources,
+            organism=organism,
+            exclude=exclude,
+        )
+        self._columns_field = var_index  # this is for consistency with BaseCatManager
         self._var_fields = var_index
         self._verify_modality(self._var_fields.keys())
         self._obs_fields = self._parse_categoricals(categoricals)
         self._modalities = set(self._var_fields.keys()) | set(self._obs_fields.keys())
-        self._using_key = using_key
         self._verbosity = verbosity
         self._obs_df_curator = None
         if "obs" in self._modalities:
-            self._obs_df_curator = DataFrameCurator(
-                df=mdata.obs,
+            self._obs_df_curator = DataFrameCatManager(
+                df=self._dataset.obs,
                 columns=Feature.name,
                 categoricals=self._obs_fields.get("obs", {}),
-                using_key=using_key,
                 verbosity=verbosity,
                 sources=self._sources.get("obs"),
                 exclude=self._exclude.get("obs"),
-                check_valid_keys=False,
-                **self._kwargs,
+                organism=organism,
             )
         self._mod_adata_curators = {
-            modality: AnnDataCurator(
-                data=mdata[modality],
+            modality: AnnDataCatManager(
+                data=self._dataset[modality],
                 var_index=var_index.get(modality),
                 categoricals=self._obs_fields.get(modality),
-                using_key=using_key,
                 verbosity=verbosity,
                 sources=self._sources.get(modality),
                 exclude=self._exclude.get(modality),
-                **self._kwargs,
+                organism=organism,
             )
             for modality in self._modalities
             if modality != "obs"
@@ -874,7 +1126,7 @@ class MuDataCurator:
         return self._obs_fields
 
     @property
-    def non_validated(self) -> dict[str, dict[str, list[str]]]:
+    def non_validated(self) -> dict[str, dict[str, list[str]]]:  # type: ignore
         """Return the non-validated features and labels."""
         if self._non_validated is None:
             raise ValidationError("Please run validate() first!")
@@ -883,15 +1135,15 @@ class MuDataCurator:
     def _verify_modality(self, modalities: Iterable[str]):
         """Verify the modality exists."""
         for modality in modalities:
-            if modality not in self._mdata.mod.keys():
+            if modality not in self._dataset.mod.keys():
                 raise ValidationError(f"modality '{modality}' does not exist!")
 
     def _parse_categoricals(self, categoricals: dict[str, FieldAttr]) -> dict:
         """Parse the categorical fields."""
-        prefixes = {f"{k}:" for k in self._mdata.mod.keys()}
+        prefixes = {f"{k}:" for k in self._dataset.mod.keys()}
         obs_fields: dict[str, dict[str, FieldAttr]] = {}
         for k, v in categoricals.items():
-            if k not in self._mdata.obs.columns:
+            if k not in self._dataset.obs.columns:
                 raise ValidationError(f"column '{k}' does not exist in mdata.obs!")
             if any(k.startswith(prefix) for prefix in prefixes):
                 modality, col = k.split(":")[0], k.split(":")[1]
@@ -904,14 +1156,11 @@ class MuDataCurator:
                 obs_fields["obs"][k] = v
         return obs_fields
 
-    def lookup(
-        self, using_key: str | None = None, public: bool = False
-    ) -> CurateLookup:
+    def lookup(self, public: bool = False) -> CurateLookup:
         """Lookup categories.
 
         Args:
-            using_key: The instance where the lookup is performed.
-                if "public", the lookup is performed on the public reference.
+            public: Perform lookup on public source ontologies.
         """
         obs_fields = {}
         for mod, fields in self._obs_fields.items():
@@ -925,27 +1174,19 @@ class MuDataCurator:
             slots={
                 **{f"{k}_var_index": v for k, v in self._var_fields.items()},
             },
-            using_key=using_key or self._using_key,
             public=public,
         )
 
+    @deprecated(new_name="is run by default")
     def add_new_from_columns(
         self,
         modality: str,
         column_names: list[str] | None = None,
-        organism: str | None = None,
         **kwargs,
     ):
-        """Update columns records."""
-        warnings.warn(
-            "`.add_new_from_columns()` is deprecated and will be removed in a future version. It's run by default during initialization.",
-            DeprecationWarning,
-            stacklevel=2,
-        )
+        pass
 
-    def add_new_from_var_index(
-        self, modality: str, organism: str | None = None, **kwargs
-    ):
+    def add_new_from_var_index(self, modality: str, **kwargs):
         """Update variable records.
 
         Args:
@@ -953,25 +1194,19 @@ class MuDataCurator:
             organism: The organism name.
             **kwargs: Additional keyword arguments to pass to create new records.
         """
-        self._kwargs.update({"organism": organism} if organism else {})
-        self._mod_adata_curators[modality].add_new_from_var_index(
-            **self._kwargs, **kwargs
-        )
+        self._mod_adata_curators[modality].add_new_from_var_index(**kwargs)
 
     def _update_registry_all(self):
         """Update all registries."""
         if self._obs_df_curator is not None:
-            self._obs_df_curator._update_registry_all(
-                validated_only=True, **self._kwargs
-            )
+            self._obs_df_curator._update_registry_all(validated_only=True)
         for _, adata_curator in self._mod_adata_curators.items():
-            adata_curator._update_registry_all(validated_only=True, **self._kwargs)
+            adata_curator._obs_df_curator._update_registry_all(validated_only=True)
 
     def add_new_from(
         self,
         key: str,
         modality: str | None = None,
-        organism: str | None = None,
         **kwargs,
     ):
         """Add validated & new categories.
@@ -984,24 +1219,17 @@ class MuDataCurator:
         """
         if len(kwargs) > 0 and key == "all":
             raise ValueError("Cannot pass additional arguments to 'all' key!")
-        self._kwargs.update({"organism": organism} if organism else {})
         modality = modality or "obs"
         if modality in self._mod_adata_curators:
             adata_curator = self._mod_adata_curators[modality]
-            adata_curator.add_new_from(key=key, **self._kwargs, **kwargs)
+            adata_curator.add_new_from(key=key, **kwargs)
         if modality == "obs":
-            self._obs_df_curator.add_new_from(key=key, **self._kwargs, **kwargs)
+            self._obs_df_curator.add_new_from(key=key, **kwargs)
 
-    def validate(self, organism: str | None = None) -> bool:
+    def validate(self) -> bool:
         """Validate categories."""
         from lamindb.core._settings import settings
 
-        self._kwargs.update({"organism": organism} if organism else {})
-        if self._using_key is not None and self._using_key != "default":
-            logger.important(
-                f"validating using registries of instance {colors.italic(self._using_key)}"
-            )
-
         # add all validated records to the current instance
         verbosity = settings.verbosity
         try:
@@ -1015,20 +1243,20 @@ class MuDataCurator:
         obs_validated = True
         if "obs" in self._modalities:
             logger.info('validating categoricals in "obs"...')
-            obs_validated &= self._obs_df_curator.validate(**self._kwargs)
+            obs_validated &= self._obs_df_curator.validate()
             self._non_validated["obs"] = self._obs_df_curator.non_validated  # type: ignore
             logger.print("")
 
         mods_validated = True
         for modality, adata_curator in self._mod_adata_curators.items():
             logger.info(f'validating categoricals in modality "{modality}"...')
-            mods_validated &= adata_curator.validate(**self._kwargs)
+            mods_validated &= adata_curator.validate()
             if len(adata_curator.non_validated) > 0:
                 self._non_validated[modality] = adata_curator.non_validated  # type: ignore
             logger.print("")
 
-        self._validated = obs_validated & mods_validated
-        return self._validated
+        self._is_validated = obs_validated & mods_validated
+        return self._is_validated
 
     def standardize(self, key: str, modality: str | None = None):
         """Replace synonyms with standardized values.
@@ -1039,6 +1267,8 @@ class MuDataCurator:
 
         Inplace modification of the dataset.
         """
+        if self._artifact is not None:
+            raise RuntimeError("can't mutate the dataset when an artifact is passed!")
         modality = modality or "obs"
         if modality in self._mod_adata_curators:
             adata_curator = self._mod_adata_curators[modality]
@@ -1046,47 +1276,6 @@ class MuDataCurator:
         if modality == "obs":
             self._obs_df_curator.standardize(key=key)
 
-    def save_artifact(
-        self,
-        description: str | None = None,
-        key: str | None = None,
-        revises: Artifact | None = None,
-        run: Run | None = None,
-    ) -> Artifact:
-        """Save the validated ``MuData`` and metadata.
-
-        Args:
-            description: A description of the ``MuData`` object.
-            key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`. Artifacts with the same key form a revision family.
-            revises: Previous version of the artifact. Triggers a revision.
-            run: The run that creates the artifact.
-
-        Returns:
-            A saved artifact record.
-        """
-        from lamindb.core._settings import settings
-
-        if not self._validated:
-            self.validate()
-            if not self._validated:
-                raise ValidationError("Dataset does not validate. Please curate.")
-        verbosity = settings.verbosity
-        try:
-            settings.verbosity = "warning"
-            self._artifact = save_artifact(
-                self._mdata,
-                description=description,
-                columns_field=self.var_index,
-                fields=self.categoricals,
-                key=key,
-                revises=revises,
-                run=run,
-                **self._kwargs,
-            )
-        finally:
-            settings.verbosity = verbosity
-        return self._artifact
-
 
 def _maybe_curation_keys_not_present(nonval_keys: list[str], name: str):
     if (n := len(nonval_keys)) > 0:
@@ -1097,8 +1286,8 @@ def _maybe_curation_keys_not_present(nonval_keys: list[str], name: str):
         )
 
 
-class SOMACurator(BaseCurator):
-    """Curation flow for ``tiledbsoma``.
+class TiledbsomaCatManager(CatManager):
+    """Curation flow for `tiledbsoma.Experiment`.
 
     See also :class:`~lamindb.Curator`.
 
@@ -1123,7 +1312,7 @@ class SOMACurator(BaseCurator):
         ...     var_index={"RNA": ("var_id", bt.Gene.symbol)},
         ...     categoricals={
         ...         "cell_type_ontology_id": bt.CellType.ontology_id,
-        ...         "donor_id": ln.ULabel.name
+        ...         "donor_id": ULabel.name
         ...     },
         ...     organism="human",
         ... )
@@ -1138,23 +1327,21 @@ class SOMACurator(BaseCurator):
         organism: str | None = None,
         sources: dict[str, Record] | None = None,
         exclude: dict[str, str | list[str]] | None = None,
-        using_key: str | None = None,
     ):
         self._obs_fields = categoricals or {}
         self._var_fields = var_index
         self._columns_field = obs_columns
         if isinstance(experiment_uri, Artifact):
-            self._experiment_uri = experiment_uri.path
+            self._dataset = experiment_uri.path
             self._artifact = experiment_uri
         else:
-            self._experiment_uri = UPath(experiment_uri)
+            self._dataset = UPath(experiment_uri)
             self._artifact = None
         self._organism = organism
-        self._using_key = using_key
         self._sources = sources or {}
         self._exclude = exclude or {}
 
-        self._validated: bool | None = False
+        self._is_validated: bool | None = False
         self._non_validated_values: dict[str, list] | None = None
         self._validated_values: dict[str, list] = {}
         # filled by _check_save_keys
@@ -1172,7 +1359,7 @@ class SOMACurator(BaseCurator):
     def _check_save_keys(self):
         from lamindb.core.storage._tiledbsoma import _open_tiledbsoma
 
-        with _open_tiledbsoma(self._experiment_uri, mode="r") as experiment:
+        with _open_tiledbsoma(self._dataset, mode="r") as experiment:
             experiment_obs = experiment.obs
             self._n_obs = len(experiment_obs)
             self._obs_pa_schema = experiment_obs.schema
@@ -1228,7 +1415,6 @@ class SOMACurator(BaseCurator):
             values=register_columns,
             field=self._columns_field,
             key="columns",
-            using_key=self._using_key,
             validated_only=False,
             organism=organism,
             source=self._sources.get("columns"),
@@ -1244,7 +1430,6 @@ class SOMACurator(BaseCurator):
                 values=additional_columns,
                 field=self._columns_field,
                 key="columns",
-                using_key=self._using_key,
                 validated_only=True,
                 organism=organism,
                 source=self._sources.get("columns"),
@@ -1257,7 +1442,7 @@ class SOMACurator(BaseCurator):
 
         validated = True
         self._non_validated_values = {}
-        with _open_tiledbsoma(self._experiment_uri, mode="r") as experiment:
+        with _open_tiledbsoma(self._dataset, mode="r") as experiment:
             for ms, (key, field) in self._var_fields.items():
                 var_ms = experiment.ms[ms].var
                 var_ms_key = f"{ms}__{key}"
@@ -1274,7 +1459,6 @@ class SOMACurator(BaseCurator):
                     values=var_ms_values,
                     field=field,
                     key=var_ms_key,
-                    using_key=self._using_key,
                     validated_only=True,
                     organism=organism,
                     source=self._sources.get(var_ms_key),
@@ -1284,7 +1468,6 @@ class SOMACurator(BaseCurator):
                     values=var_ms_values,
                     field=field,
                     key=var_ms_key,
-                    using_key=self._using_key,
                     organism=organism,
                     source=self._sources.get(var_ms_key),
                     exclude=self._exclude.get(var_ms_key),
@@ -1310,7 +1493,6 @@ class SOMACurator(BaseCurator):
                     values=values,
                     field=field,
                     key=key,
-                    using_key=self._using_key,
                     validated_only=True,
                     organism=organism,
                     source=self._sources.get(key),
@@ -1320,7 +1502,6 @@ class SOMACurator(BaseCurator):
                     values=values,
                     field=field,
                     key=key,
-                    using_key=self._using_key,
                     organism=organism,
                     source=self._sources.get(key),
                     exclude=self._exclude.get(key),
@@ -1330,8 +1511,8 @@ class SOMACurator(BaseCurator):
                     self._non_validated_values[key] = non_val
                 else:
                     self._validated_values[key] = values
-        self._validated = validated
-        return self._validated
+        self._is_validated = validated
+        return self._is_validated
 
     def _non_validated_values_field(self, key: str) -> tuple[list, FieldAttr]:
         assert self._non_validated_values is not None  # noqa: S101
@@ -1346,7 +1527,7 @@ class SOMACurator(BaseCurator):
         values = self._non_validated_values.get(key, [])
         return values, field
 
-    def add_new_from(self, key: str) -> None:
+    def add_new_from(self, key: str, **kwargs) -> None:
         """Add validated & new categories.
 
         Args:
@@ -1378,11 +1559,11 @@ class SOMACurator(BaseCurator):
                 values=values,
                 field=field,
                 key=k,
-                using_key=self._using_key,
                 validated_only=False,
                 organism=organism,
                 source=self._sources.get(k),
                 exclude=self._exclude.get(k),
+                **kwargs,
             )
             # update non-validated values list but keep the key there
             # it will be removed by .validate()
@@ -1405,19 +1586,15 @@ class SOMACurator(BaseCurator):
         """Return the obs fields to validate against."""
         return self._obs_fields
 
-    def lookup(
-        self, using_key: str | None = None, public: bool = False
-    ) -> CurateLookup:
+    def lookup(self, public: bool = False) -> CurateLookup:
         """Lookup categories.
 
         Args:
-            using_key: The instance where the lookup is performed.
-                if "public", the lookup is performed on the public reference.
+            public: If "public", the lookup is performed on the public reference.
         """
         return CurateLookup(
             categoricals=self._obs_fields,
             slots={"columns": self._columns_field, **self._var_fields_flat},
-            using_key=using_key or self._using_key,
             public=public,
         )
 
@@ -1462,7 +1639,6 @@ class SOMACurator(BaseCurator):
             syn_mapper = standardize_categories(
                 values=values,
                 field=field,
-                using_key=self._using_key,
                 source=self._sources.get(k),
                 organism=organism,
             )
@@ -1471,7 +1647,7 @@ class SOMACurator(BaseCurator):
 
             from lamindb.core.storage._tiledbsoma import _open_tiledbsoma
 
-            with _open_tiledbsoma(self._experiment_uri, mode="r") as experiment:
+            with _open_tiledbsoma(self._dataset, mode="r") as experiment:
                 value_filter = f"{slot_key} in {list(syn_mapper.keys())}"
                 table = slot(experiment).read(value_filter=value_filter).concat()
 
@@ -1484,7 +1660,7 @@ class SOMACurator(BaseCurator):
                 lambda val: syn_mapper.get(val, val)  # noqa
             )
             # write the mapped values
-            with _open_tiledbsoma(self._experiment_uri, mode="w") as experiment:
+            with _open_tiledbsoma(self._dataset, mode="w") as experiment:
                 slot(experiment).write(pa.Table.from_pandas(df, schema=table.schema))
             # update non_validated dict
             non_val_k = [
@@ -1502,8 +1678,9 @@ class SOMACurator(BaseCurator):
 
     def save_artifact(
         self,
-        description: str | None = None,
+        *,
         key: str | None = None,
+        description: str | None = None,
         revises: Artifact | None = None,
         run: Run | None = None,
     ) -> Artifact:
@@ -1512,7 +1689,7 @@ class SOMACurator(BaseCurator):
         Args:
             description: A description of the ``tiledbsoma`` store.
             key: A path-like key to reference artifact in default storage,
-                e.g., `"myfolder/mystore.tiledbsoma"`. Artifacts with the same key form a revision family.
+                e.g., `"myfolder/mystore.tiledbsoma"`. Artifacts with the same key form a version family.
             revises: Previous version of the artifact. Triggers a revision.
             run: The run that creates the artifact.
 
@@ -1521,14 +1698,14 @@ class SOMACurator(BaseCurator):
         """
         from lamindb.core._data import add_labels
 
-        if not self._validated:
+        if not self._is_validated:
             self.validate()
-            if not self._validated:
+            if not self._is_validated:
                 raise ValidationError("Dataset does not validate. Please curate.")
 
         if self._artifact is None:
             artifact = Artifact(
-                self._experiment_uri,
+                self._dataset,
                 description=description,
                 key=key,
                 revises=revises,
@@ -1540,7 +1717,7 @@ class SOMACurator(BaseCurator):
         else:
             artifact = self._artifact
 
-        _schemas_m2m = {}
+        feature_sets = {}
         if len(self._obs_fields) > 0:
             organism = check_registry_organism(
                 self._columns_field.field.model, self._organism
@@ -1550,7 +1727,7 @@ class SOMACurator(BaseCurator):
                 empty_dict, schema=self._obs_pa_schema
             ).to_pandas()
             # in parallel to https://github.com/laminlabs/lamindb/blob/2a1709990b5736b480c6de49c0ada47fafc8b18d/lamindb/core/_feature_manager.py#L549-L554
-            _schemas_m2m["obs"] = Schema.from_df(
+            feature_sets["obs"] = Schema.from_df(
                 df=mock_df,
                 field=self._columns_field,
                 mute=True,
@@ -1561,238 +1738,1370 @@ class SOMACurator(BaseCurator):
             organism = check_registry_organism(
                 var_field.field.model, self._organism
             ).get("organism")
-            _schemas_m2m[f"{ms}__var"] = Schema.from_values(
+            feature_sets[f"{ms}__var"] = Schema.from_values(
                 values=self._validated_values[f"{ms}__{var_key}"],
                 field=var_field,
                 organism=organism,
                 raise_validation_error=False,
             )
-        artifact._staged__schemas_m2m = _schemas_m2m
+        artifact._staged_feature_sets = feature_sets
+
+        feature_ref_is_name = _ref_is_name(self._columns_field)
+        features = Feature.lookup().dict()
+        for key, field in self._obs_fields.items():
+            feature = features.get(key)
+            registry = field.field.model
+            organism = check_registry_organism(field.field.model, self._organism).get(
+                "organism"
+            )
+            labels = registry.from_values(
+                values=self._validated_values[key], field=field, organism=organism
+            )
+            if len(labels) == 0:
+                continue
+            if hasattr(registry, "_name_field"):
+                label_ref_is_name = field.field.name == registry._name_field
+                add_labels(
+                    artifact,
+                    records=labels,
+                    feature=feature,
+                    feature_ref_is_name=feature_ref_is_name,
+                    label_ref_is_name=label_ref_is_name,
+                    from_curator=True,
+                )
+
+        return artifact.save()
+
+
+class SpatialDataCatManager(CatManager):
+    """Curation flow for a ``Spatialdata`` object.
+
+    See also :class:`~lamindb.Curator`.
+
+    Note that if genes or other measurements are removed from the SpatialData object,
+    the object should be recreated.
+
+    In the following docstring, an accessor refers to either a ``.table`` key or the ``sample_metadata_key``.
+
+    Args:
+        sdata: The SpatialData object to curate.
+        var_index: A dictionary mapping table keys to the ``.var`` indices.
+        categoricals: A nested dictionary mapping an accessor to dictionaries that map columns to a registry field.
+
+        organism: The organism name.
+        sources: A dictionary mapping an accessor to dictionaries that map columns to Source records.
+        exclude: A dictionary mapping an accessor to dictionaries of column names to values to exclude from validation.
+            When specific :class:`~bionty.Source` instances are pinned and may lack default values (e.g., "unknown" or "na"),
+            using the exclude parameter ensures they are not validated.
+        verbosity: The verbosity level of the logger.
+        sample_metadata_key: The key in ``.attrs`` that stores the sample level metadata.
+
+    Examples:
+        >>> import bionty as bt
+        >>> curator = SpatialDataCatManager(
+        ...     sdata,
+        ...     var_index={
+        ...         "table_1": bt.Gene.ensembl_gene_id,
+        ...     },
+        ...     categoricals={
+        ...         "table1":
+        ...             {"cell_type_ontology_id": bt.CellType.ontology_id, "donor_id": ULabel.name},
+        ...         "sample":
+        ...             {"experimental_factor": bt.ExperimentalFactor.name},
+        ...     },
+        ...     organism="human",
+        ... )
+    """
+
+    def __init__(
+        self,
+        sdata: Any,
+        var_index: dict[str, FieldAttr],
+        categoricals: dict[str, dict[str, FieldAttr]] | None = None,
+        verbosity: str = "hint",
+        organism: str | None = None,
+        sources: dict[str, dict[str, Record]] | None = None,
+        exclude: dict[str, dict] | None = None,
+        *,
+        sample_metadata_key: str | None = "sample",
+    ) -> None:
+        super().__init__(
+            dataset=sdata,
+            categoricals={},
+            sources=sources,
+            organism=organism,
+            exclude=exclude,
+        )
+        if isinstance(sdata, Artifact):
+            # TODO: load() doesn't yet work
+            self._sdata = sdata.load()
+        else:
+            self._sdata = self._dataset
+        self._sample_metadata_key = sample_metadata_key
+        self._var_fields = var_index
+        self._verify_accessor_exists(self._var_fields.keys())
+        self._categoricals = categoricals
+        self._table_keys = set(self._var_fields.keys()) | set(
+            self._categoricals.keys() - {self._sample_metadata_key}
+        )
+        self._verbosity = verbosity
+        self._sample_df_curator = None
+        if self._sample_metadata_key is not None:
+            self._sample_metadata = self._sdata.get_attrs(
+                key=self._sample_metadata_key, return_as="df", flatten=True
+            )
+        self._is_validated = False
+
+        # Check validity of keys in categoricals
+        nonval_keys = []
+        for accessor, accessor_categoricals in self._categoricals.items():
+            if (
+                accessor == self._sample_metadata_key
+                and self._sample_metadata is not None
+            ):
+                for key in accessor_categoricals.keys():
+                    if key not in self._sample_metadata.columns:
+                        nonval_keys.append(key)
+            else:
+                for key in accessor_categoricals.keys():
+                    if key not in self._sdata[accessor].obs.columns:
+                        nonval_keys.append(key)
+
+        _maybe_curation_keys_not_present(nonval_keys, "categoricals")
+
+        # check validity of keys in sources and exclude
+        for name, dct in (("sources", self._sources), ("exclude", self._exclude)):
+            nonval_keys = []
+            for accessor, accessor_sources in dct.items():
+                if (
+                    accessor == self._sample_metadata_key
+                    and self._sample_metadata is not None
+                ):
+                    columns = self._sample_metadata.columns
+                elif accessor != self._sample_metadata_key:
+                    columns = self._sdata[accessor].obs.columns
+                else:
+                    continue
+                for key in accessor_sources:
+                    if key not in columns:
+                        nonval_keys.append(key)
+            _maybe_curation_keys_not_present(nonval_keys, name)
+
+        # Set up sample level metadata and table Curator objects
+        if (
+            self._sample_metadata_key is not None
+            and self._sample_metadata_key in self._categoricals
+        ):
+            self._sample_df_curator = DataFrameCatManager(
+                df=self._sample_metadata,
+                columns=Feature.name,
+                categoricals=self._categoricals.get(self._sample_metadata_key, {}),
+                verbosity=verbosity,
+                sources=self._sources.get(self._sample_metadata_key),
+                exclude=self._exclude.get(self._sample_metadata_key),
+                organism=organism,
+            )
+        self._table_adata_curators = {
+            table: AnnDataCatManager(
+                data=self._sdata[table],
+                var_index=var_index.get(table),
+                categoricals=self._categoricals.get(table),
+                verbosity=verbosity,
+                sources=self._sources.get(table),
+                exclude=self._exclude.get(table),
+                organism=organism,
+            )
+            for table in self._table_keys
+        }
+
+        self._non_validated = None
+
+    @property
+    def var_index(self) -> FieldAttr:
+        """Return the registry fields to validate variables indices against."""
+        return self._var_fields
+
+    @property
+    def categoricals(self) -> dict[str, dict[str, FieldAttr]]:
+        """Return the categorical keys and fields to validate against."""
+        return self._categoricals
+
+    @property
+    def non_validated(self) -> dict[str, dict[str, list[str]]]:  # type: ignore
+        """Return the non-validated features and labels."""
+        if self._non_validated is None:
+            raise ValidationError("Please run validate() first!")
+        return self._non_validated
+
+    def _verify_accessor_exists(self, accessors: Iterable[str]) -> None:
+        """Verify that the accessors exist (either a valid table or in attrs)."""
+        for acc in accessors:
+            is_present = False
+            try:
+                self._sdata.get_attrs(key=acc)
+                is_present = True
+            except KeyError:
+                if acc in self._sdata.tables.keys():
+                    is_present = True
+            if not is_present:
+                raise ValidationError(f"Accessor '{acc}' does not exist!")
+
+    def lookup(self, public: bool = False) -> CurateLookup:
+        """Look up categories.
+
+        Args:
+            public: Whether the lookup is performed on the public reference.
+        """
+        cat_values_dict = list(self.categoricals.values())[0]
+        return CurateLookup(
+            categoricals=cat_values_dict,
+            slots={"accessors": cat_values_dict.keys()},
+            public=public,
+        )
+
+    def _update_registry_all(self) -> None:
+        """Saves labels of all features for sample and table metadata."""
+        if self._sample_df_curator is not None:
+            self._sample_df_curator._update_registry_all(
+                validated_only=True,
+            )
+        for _, adata_curator in self._table_adata_curators.items():
+            adata_curator._obs_df_curator._update_registry_all(
+                validated_only=True,
+            )
+
+    def add_new_from_var_index(self, table: str, **kwargs) -> None:
+        """Save new values from ``.var.index`` of table.
+
+        Args:
+            table: The table key.
+            organism: The organism name.
+            **kwargs: Additional keyword arguments to pass to create new records.
+        """
+        if self._non_validated is None:
+            raise ValidationError("Run .validate() first.")
+        self._table_adata_curators[table].add_new_from_var_index(**kwargs)
+        if table in self.non_validated.keys():
+            if "var_index" in self._non_validated[table]:
+                self._non_validated[table].pop("var_index")
+
+            if len(self.non_validated[table].values()) == 0:
+                self.non_validated.pop(table)
+
+    def add_new_from(
+        self,
+        key: str,
+        accessor: str | None = None,
+        **kwargs,
+    ) -> None:
+        """Save new values of categorical from sample level metadata or table.
+
+        Args:
+            key: The key referencing the slot in the DataFrame.
+            accessor: The accessor key such as 'sample' or 'table x'.
+            organism: The organism name.
+            **kwargs: Additional keyword arguments to pass to create new records.
+        """
+        if self._non_validated is None:
+            raise ValidationError("Run .validate() first.")
+
+        if len(kwargs) > 0 and key == "all":
+            raise ValueError("Cannot pass additional arguments to 'all' key!")
+
+        if accessor not in self.categoricals:
+            raise ValueError(
+                f"Accessor {accessor} is not in 'categoricals'. Include it when creating the SpatialDataCatManager."
+            )
+
+        if accessor in self._table_adata_curators:
+            adata_curator = self._table_adata_curators[accessor]
+            adata_curator.add_new_from(key=key, **kwargs)
+        if accessor == self._sample_metadata_key:
+            self._sample_df_curator.add_new_from(key=key, **kwargs)
+
+        if accessor in self.non_validated.keys():
+            if len(self.non_validated[accessor].values()) == 0:
+                self.non_validated.pop(accessor)
+
+    def standardize(self, key: str, accessor: str | None = None) -> None:
+        """Replace synonyms with canonical values.
+
+        Modifies the dataset inplace.
+
+        Args:
+            key: The key referencing the slot in the table or sample metadata.
+            accessor: The accessor key such as 'sample_key' or 'table_key'.
+        """
+        if len(self.non_validated) == 0:
+            logger.warning("values are already standardized")
+            return
+        if self._artifact is not None:
+            raise RuntimeError("can't mutate the dataset when an artifact is passed!")
+
+        if accessor == self._sample_metadata_key:
+            if key not in self._sample_metadata.columns:
+                raise ValueError(f"key '{key}' not present in '{accessor}'!")
+        else:
+            if (
+                key == "var_index" and self._sdata.tables[accessor].var.index is None
+            ) or (
+                key != "var_index"
+                and key not in self._sdata.tables[accessor].obs.columns
+            ):
+                raise ValueError(f"key '{key}' not present in '{accessor}'!")
+
+        if accessor in self._table_adata_curators.keys():
+            adata_curator = self._table_adata_curators[accessor]
+            adata_curator.standardize(key)
+        if accessor == self._sample_metadata_key:
+            self._sample_df_curator.standardize(key)
+
+        if len(self.non_validated[accessor].values()) == 0:
+            self.non_validated.pop(accessor)
+
+    def validate(self) -> bool:
+        """Validate variables and categorical observations.
+
+        This method also registers the validated records in the current instance:
+        - from public sources
+
+        Args:
+            organism: The organism name.
+
+        Returns:
+            Whether the SpatialData object is validated.
+        """
+        from lamindb.core._settings import settings
+
+        # add all validated records to the current instance
+        verbosity = settings.verbosity
+        try:
+            settings.verbosity = "error"
+            self._update_registry_all()
+        finally:
+            settings.verbosity = verbosity
+
+        self._non_validated = {}  # type: ignore
+
+        sample_validated = True
+        if self._sample_df_curator:
+            logger.info(f"validating categoricals of '{self._sample_metadata_key}' ...")
+            sample_validated &= self._sample_df_curator.validate()
+            if len(self._sample_df_curator.non_validated) > 0:
+                self._non_validated["sample"] = self._sample_df_curator.non_validated  # type: ignore
+            logger.print("")
+
+        mods_validated = True
+        for table, adata_curator in self._table_adata_curators.items():
+            logger.info(f"validating categoricals of table '{table}' ...")
+            mods_validated &= adata_curator.validate()
+            if len(adata_curator.non_validated) > 0:
+                self._non_validated[table] = adata_curator.non_validated  # type: ignore
+            logger.print("")
+
+        self._is_validated = sample_validated & mods_validated
+        return self._is_validated
+
+    def save_artifact(
+        self,
+        *,
+        key: str | None = None,
+        description: str | None = None,
+        revises: Artifact | None = None,
+        run: Run | None = None,
+    ) -> Artifact:
+        if not self._is_validated:
+            self.validate()
+            if not self._is_validated:
+                raise ValidationError("Dataset does not validate. Please curate.")
+
+        verbosity = settings.verbosity
+        try:
+            settings.verbosity = "warning"
+
+            if self._artifact is None:
+                # Write the SpatialData object to a random path in tmp directory
+                # The Artifact constructor will move it to the cache
+                write_path = (
+                    f"{settings.cache_dir}/{random.randint(10**7, 10**8 - 1)}.zarr"
+                )
+                self._sdata.write(write_path)
+
+                # Create the Artifact and associate Artifact metadata
+                self._artifact = Artifact(
+                    write_path,
+                    description=description,
+                    key=key,
+                    revises=revises,
+                    run=run,
+                )
+                # According to Tim it is not easy to calculate the number of observations.
+                # We would have to write custom code to iterate over labels (which might not even exist at that point)
+                self._artifact.otype = "spatialdata"
+                self._artifact.save()
+
+            # Link schemas
+            feature_kwargs = check_registry_organism(
+                (list(self._var_fields.values())[0].field.model),
+                self._organism,
+            )
+
+            def _add_set_from_spatialdata(
+                host: Artifact | Collection | Run,
+                var_fields: dict[str, FieldAttr],
+                obs_fields: dict[str, FieldAttr] = None,
+                mute: bool = False,
+                organism: str | Record | None = None,
+            ):
+                """Add Schemas from SpatialData."""
+                if obs_fields is None:
+                    obs_fields = {}
+                assert host.otype == "spatialdata"  # noqa: S101
+
+                feature_sets = {}
+
+                # sample features
+                sample_features = Feature.from_values(self._sample_metadata.columns)  # type: ignore
+                if len(sample_features) > 0:
+                    feature_sets[self._sample_metadata_key] = Schema(
+                        features=sample_features
+                    )
+
+                # table features
+                for table, field in var_fields.items():
+                    table_fs = parse_staged_feature_sets_from_anndata(
+                        self._sdata[table],
+                        var_field=field,
+                        obs_field=obs_fields.get(table, Feature.name),
+                        mute=mute,
+                        organism=organism,
+                    )
+                    for k, v in table_fs.items():
+                        feature_sets[f"['{table}'].{k}"] = v
+
+                def _unify_staged_feature_sets_by_hash(
+                    feature_sets: MutableMapping[str, Schema],
+                ):
+                    unique_values: dict[str, Any] = {}
+
+                    for key, value in feature_sets.items():
+                        value_hash = (
+                            value.hash
+                        )  # Assuming each value has a .hash attribute
+                        if value_hash in unique_values:
+                            feature_sets[key] = unique_values[value_hash]
+                        else:
+                            unique_values[value_hash] = value
+
+                    return feature_sets
+
+                # link feature sets
+                host._staged_feature_sets = _unify_staged_feature_sets_by_hash(
+                    feature_sets
+                )
+                host.save()
+
+            _add_set_from_spatialdata(
+                self._artifact, var_fields=self._var_fields, **feature_kwargs
+            )
+
+            # Link labels
+            def _add_labels_from_spatialdata(
+                data,
+                artifact: Artifact,
+                fields: dict[str, FieldAttr],
+                feature_ref_is_name: bool | None = None,
+            ):
+                """Add Labels from SpatialData."""
+                features = Feature.lookup().dict()
+                for key, field in fields.items():
+                    feature = features.get(key)
+                    registry = field.field.model
+                    filter_kwargs = check_registry_organism(registry, self._organism)
+                    filter_kwargs_current = get_current_filter_kwargs(
+                        registry, filter_kwargs
+                    )
+                    df = data if isinstance(data, pd.DataFrame) else data.obs
+                    labels = registry.from_values(
+                        df[key],
+                        field=field,
+                        **filter_kwargs_current,
+                    )
+                    if len(labels) == 0:
+                        continue
+
+                    label_ref_is_name = None
+                    if hasattr(registry, "_name_field"):
+                        label_ref_is_name = field.field.name == registry._name_field
+                    add_labels(
+                        artifact,
+                        records=labels,
+                        feature=feature,
+                        feature_ref_is_name=feature_ref_is_name,
+                        label_ref_is_name=label_ref_is_name,
+                        from_curator=True,
+                    )
+
+            for accessor, accessor_fields in self._categoricals.items():
+                column_field = self._var_fields.get(accessor)
+                if accessor == self._sample_metadata_key:
+                    _add_labels_from_spatialdata(
+                        self._sample_metadata,
+                        self._artifact,
+                        accessor_fields,
+                        feature_ref_is_name=(
+                            None if column_field is None else _ref_is_name(column_field)
+                        ),
+                    )
+                else:
+                    _add_labels_from_spatialdata(
+                        self._sdata.tables[accessor],
+                        self._artifact,
+                        accessor_fields,
+                        feature_ref_is_name=(
+                            None if column_field is None else _ref_is_name(column_field)
+                        ),
+                    )
+
+        finally:
+            settings.verbosity = verbosity
+
+        slug = ln_setup.settings.instance.slug
+        if ln_setup.settings.instance.is_remote:  # pragma: no cover
+            logger.important(
+                f"go to https://lamin.ai/{slug}/artifact/{self._artifact.uid}"
+            )
+
+        return self._artifact
+
+
+def _restrict_obs_fields(
+    obs: pd.DataFrame, obs_fields: dict[str, FieldAttr]
+) -> dict[str, str]:
+    """Restrict the obs fields to name return only available obs fields.
+
+    To simplify the curation, we only validate against either name or ontology_id.
+    If both are available, we validate against ontology_id.
+    If none are available, we validate against name.
+    """
+    obs_fields_unique = {k: v for k, v in obs_fields.items() if k in obs.columns}
+    for name, field in obs_fields.items():
+        if name.endswith("_ontology_term_id"):
+            continue
+        # if both the ontology id and the name are present, only validate on the ontology_id
+        if name in obs.columns and f"{name}_ontology_term_id" in obs.columns:
+            obs_fields_unique.pop(name)
+        # if the neither name nor ontology id are present, validate on the name
+        # this will raise error downstream, we just use name to be more readable
+        if name not in obs.columns and f"{name}_ontology_term_id" not in obs.columns:
+            obs_fields_unique[name] = field
+
+    # Only retain obs_fields_unique that have keys in adata.obs.columns
+    available_obs_fields = {
+        k: v for k, v in obs_fields_unique.items() if k in obs.columns
+    }
+
+    return available_obs_fields
+
+
+def _add_defaults_to_obs(
+    obs: pd.DataFrame,
+    defaults: dict[str, str],
+) -> None:
+    """Add default columns and values to obs DataFrame."""
+    added_defaults: dict = {}
+    for name, default in defaults.items():
+        if name not in obs.columns and f"{name}_ontology_term_id" not in obs.columns:
+            obs[name] = default
+            added_defaults[name] = default
+            logger.important(
+                f"added default value '{default}' to the adata.obs['{name}']"
+            )
+
+
+class CellxGeneAnnDataCatManager(AnnDataCatManager):
+    """Annotation flow of AnnData based on CELLxGENE schema."""
+
+    _controls_were_created: bool | None = None
+
+    def __init__(
+        self,
+        adata: ad.AnnData | UPathStr,
+        categoricals: dict[str, FieldAttr] | None = None,
+        organism: Literal["human", "mouse"] = "human",
+        *,
+        defaults: dict[str, str] = None,
+        extra_sources: dict[str, Record] = None,
+        schema_version: Literal["4.0.0", "5.0.0", "5.1.0"] = "5.1.0",
+        verbosity: str = "hint",
+    ) -> None:
+        """CELLxGENE schema curator.
+
+        Args:
+            adata: Path to or AnnData object to curate against the CELLxGENE schema.
+            categoricals: A dictionary mapping ``.obs.columns`` to a registry field.
+                The CELLxGENE Curator maps against the required CELLxGENE fields by default.
+            organism: The organism name. CELLxGENE restricts it to 'human' and 'mouse'.
+            defaults: Default values that are set if columns or column values are missing.
+            extra_sources: A dictionary mapping ``.obs.columns`` to Source records.
+                These extra sources are joined with the CELLxGENE fixed sources.
+                Use this parameter when subclassing.
+            exclude: A dictionary mapping column names to values to exclude.
+            schema_version: The CELLxGENE schema version to curate against.
+            verbosity: The verbosity level.
+
+        """
+        import bionty as bt
+
+        CellxGeneAnnDataCatManager._init_categoricals_additional_values()
+
+        var_index: FieldAttr = bt.Gene.ensembl_gene_id
+
+        if categoricals is None:
+            categoricals = CellxGeneAnnDataCatManager._get_categoricals()
+
+        self.organism = organism
+
+        VALID_SCHEMA_VERSIONS = {"4.0.0", "5.0.0", "5.1.0"}
+        if schema_version not in VALID_SCHEMA_VERSIONS:
+            valid_versions = ", ".join(sorted(VALID_SCHEMA_VERSIONS))
+            raise ValueError(
+                f"Invalid schema_version: {schema_version}. "
+                f"Valid versions are: {valid_versions}"
+            )
+        self.schema_version = schema_version
+        self.schema_reference = f"https://github.com/chanzuckerberg/single-cell-curation/blob/main/schema/{schema_version}/schema.md"
+        with resources.path(
+            "lamindb.curators._cellxgene_schemas", "schema_versions.yml"
+        ) as schema_versions_path:
+            self._pinned_ontologies = _read_schema_versions(schema_versions_path)[
+                self.schema_version
+            ]
+
+        # Fetch AnnData obs to be able to set defaults and get sources
+        if isinstance(adata, ad.AnnData):
+            self._adata_obs = adata.obs
+        else:
+            self._adata_obs = backed_access(upath.create_path(adata)).obs  # type: ignore
+
+        # Add defaults first to ensure that we fetch valid sources
+        if defaults:
+            _add_defaults_to_obs(self._adata_obs, defaults)
+
+        self.sources = self._create_sources(self._adata_obs)
+        self.sources = {
+            entity: source
+            for entity, source in self.sources.items()
+            if source is not None
+        }
+
+        # These sources are not a part of the cellxgene schema but rather passed through.
+        # This is useful when other Curators extend the CELLxGENE curator
+        if extra_sources:
+            self.sources = self.sources | extra_sources
+
+        # Exclude default values from validation because they are not available in the pinned sources
+        exclude_keys = {
+            entity: default
+            for entity, default in CellxGeneAnnDataCatManager._get_categoricals_defaults().items()
+            if entity in self._adata_obs.columns  # type: ignore
+        }
+
+        super().__init__(
+            data=adata,
+            var_index=var_index,
+            categoricals=_restrict_obs_fields(self._adata_obs, categoricals),
+            verbosity=verbosity,
+            organism=organism,
+            sources=self.sources,
+            exclude=exclude_keys,
+        )
+
+    @classmethod
+    def _init_categoricals_additional_values(cls) -> None:
+        import bionty as bt
+
+        import lamindb as ln
+
+        # Note: if you add another control below, be mindful to change the if condition that
+        # triggers whether creating these records is re-considered
+        if cls._controls_were_created is None:
+            cls._controls_were_created = (
+                ln.ULabel.filter(name="SuspensionType", is_type=True).one_or_none()
+                is not None
+            )
+        if not cls._controls_were_created:
+            logger.important("Creating control labels in the CellxGene schema.")
+            bt.CellType(
+                ontology_id="unknown",
+                name="unknown",
+                description="From CellxGene schema.",
+            ).save()
+            pato = bt.Source.filter(name="pato", version="2024-03-28").one()
+            normal = bt.Phenotype.from_source(ontology_id="PATO:0000461", source=pato)
+            bt.Disease(
+                uid=normal.uid,
+                name=normal.name,
+                ontology_id=normal.ontology_id,
+                description=normal.description,
+                source=normal.source,
+            ).save()
+            bt.Ethnicity(
+                ontology_id="na", name="na", description="From CellxGene schema."
+            ).save()
+            bt.Ethnicity(
+                ontology_id="unknown",
+                name="unknown",
+                description="From CellxGene schema.",
+            ).save()
+            bt.DevelopmentalStage(
+                ontology_id="unknown",
+                name="unknown",
+                description="From CellxGene schema.",
+            ).save()
+            bt.Phenotype(
+                ontology_id="unknown",
+                name="unknown",
+                description="From CellxGene schema.",
+            ).save()
+
+            tissue_type = ln.ULabel(
+                name="TissueType",
+                is_type=True,
+                description='From CellxGene schema. Is "tissue", "organoid", or "cell culture".',
+            ).save()
+            ln.ULabel(
+                name="tissue", type=tissue_type, description="From CellxGene schema."
+            ).save()
+            ln.ULabel(
+                name="organoid", type=tissue_type, description="From CellxGene schema."
+            ).save()
+            ln.ULabel(
+                name="cell culture",
+                type=tissue_type,
+                description="From CellxGene schema.",
+            ).save()
+
+            suspension_type = ln.ULabel(
+                name="SuspensionType",
+                is_type=True,
+                description='From CellxGene schema. This MUST be "cell", "nucleus", or "na".',
+            ).save()
+            ln.ULabel(
+                name="cell", type=suspension_type, description="From CellxGene schema."
+            ).save()
+            ln.ULabel(
+                name="nucleus",
+                type=suspension_type,
+                description="From CellxGene schema.",
+            ).save()
+            ln.ULabel(name="na", type=suspension_type).save()
+
+    @classmethod
+    def _get_categoricals(cls) -> dict[str, FieldAttr]:
+        import bionty as bt
+
+        return {
+            "assay": bt.ExperimentalFactor.name,
+            "assay_ontology_term_id": bt.ExperimentalFactor.ontology_id,
+            "cell_type": bt.CellType.name,
+            "cell_type_ontology_term_id": bt.CellType.ontology_id,
+            "development_stage": bt.DevelopmentalStage.name,
+            "development_stage_ontology_term_id": bt.DevelopmentalStage.ontology_id,
+            "disease": bt.Disease.name,
+            "disease_ontology_term_id": bt.Disease.ontology_id,
+            # "donor_id": "str",  via pandera
+            "self_reported_ethnicity": bt.Ethnicity.name,
+            "self_reported_ethnicity_ontology_term_id": bt.Ethnicity.ontology_id,
+            "sex": bt.Phenotype.name,
+            "sex_ontology_term_id": bt.Phenotype.ontology_id,
+            "suspension_type": ULabel.name,
+            "tissue": bt.Tissue.name,
+            "tissue_ontology_term_id": bt.Tissue.ontology_id,
+            "tissue_type": ULabel.name,
+            "organism": bt.Organism.name,
+            "organism_ontology_term_id": bt.Organism.ontology_id,
+        }
+
+    @classmethod
+    def _get_categoricals_defaults(cls) -> dict[str, str]:
+        return {
+            "cell_type": "unknown",
+            "development_stage": "unknown",
+            "disease": "normal",
+            "donor_id": "unknown",
+            "self_reported_ethnicity": "unknown",
+            "sex": "unknown",
+            "suspension_type": "cell",
+            "tissue_type": "tissue",
+        }
+
+    @property
+    def pinned_ontologies(self) -> pd.DataFrame:
+        return self._pinned_ontologies
+
+    @property
+    def adata(self) -> AnnData:
+        return self._adata
+
+    def _create_sources(self, obs: pd.DataFrame) -> dict[str, Record]:
+        """Creates a sources dictionary that can be passed to AnnDataCatManager."""
+        import bionty as bt
+
+        # fmt: off
+        def _fetch_bionty_source(
+            entity: str, organism: str, source: str
+        ) -> bt.Source | None:
+            """Fetch the Bionty source of the pinned ontology.
+
+            Returns None if the source does not exist.
+            """
+            version = self._pinned_ontologies.loc[(self._pinned_ontologies.index == entity) &
+                                                  (self._pinned_ontologies["organism"] == organism) &
+                                                  (self._pinned_ontologies["source"] == source), "version"].iloc[0]
+            return bt.Source.filter(organism=organism, entity=f"bionty.{entity}", version=version).first()
+
+        entity_mapping = {
+             "var_index": ("Gene", self.organism, "ensembl"),
+             "cell_type": ("CellType", "all", "cl"),
+             "assay": ("ExperimentalFactor", "all", "efo"),
+             "self_reported_ethnicity": ("Ethnicity", self.organism, "hancestro"),
+             "development_stage": ("DevelopmentalStage", self.organism, "hsapdv" if self.organism == "human" else "mmusdv"),
+             "disease": ("Disease", "all", "mondo"),
+             # "organism": ("Organism", "vertebrates", "ensembl"),
+             "sex": ("Phenotype", "all", "pato"),
+             "tissue": ("Tissue", "all", "uberon"),
+        }
+        # fmt: on
+
+        # Retain var_index and one of 'entity'/'entity_ontology_term_id' that is present in obs
+        entity_to_sources = {
+            entity: _fetch_bionty_source(*params)
+            for entity, params in entity_mapping.items()
+            if entity in obs.columns
+            or (f"{entity}_ontology_term_id" in obs.columns and entity != "var_index")
+            or entity == "var_index"
+        }
+
+        return entity_to_sources
+
+    def _convert_name_to_ontology_id(self, values: pd.Series, field: FieldAttr):
+        """Converts a column that stores a name into a column that stores the ontology id.
+
+        cellxgene expects the obs columns to be {entity}_ontology_id columns and disallows {entity} columns.
+        """
+        field_name = field.field.name
+        assert field_name == "name"  # noqa: S101
+        cols = ["name", "ontology_id"]
+        registry = field.field.model
+
+        if hasattr(registry, "ontology_id"):
+            validated_records = registry.filter(**{f"{field_name}__in": values})
+            mapper = (
+                pd.DataFrame(validated_records.values_list(*cols))
+                .set_index(0)
+                .to_dict()[1]
+            )
+            return values.map(mapper)
+
+    def validate(self) -> bool:  # type: ignore
+        """Validates the AnnData object against most cellxgene requirements."""
+        # Verify that all required obs columns are present
+        missing_obs_fields = [
+            name
+            for name in CellxGeneAnnDataCatManager._get_categoricals_defaults().keys()
+            if name not in self._adata.obs.columns
+            and f"{name}_ontology_term_id" not in self._adata.obs.columns
+        ]
+        if len(missing_obs_fields) > 0:
+            missing_obs_fields_str = ", ".join(list(missing_obs_fields))
+            logger.error(f"missing required obs columns {missing_obs_fields_str}")
+            logger.info(
+                "consider initializing a Curate object like 'Curate(adata, defaults=cxg.CellxGeneAnnDataCatManager._get_categoricals_defaults())'"
+                "to automatically add these columns with default values."
+            )
+            return False
+
+        # Verify that no cellxgene reserved names are present
+        reserved_names = {
+            "ethnicity",
+            "ethnicity_ontology_term_id",
+            "X_normalization",
+            "default_field",
+            "layer_descriptions",
+            "tags",
+            "versions",
+            "contributors",
+            "preprint_doi",
+            "project_description",
+            "project_links",
+            "project_name",
+            "publication_doi",
+        }
+        matched_columns = [
+            column for column in self._adata.obs.columns if column in reserved_names
+        ]
+        if len(matched_columns) > 0:
+            raise ValueError(
+                f"AnnData object must not contain obs columns {matched_columns} which are"
+                " reserved from previous schema versions."
+            )
 
-        feature_ref_is_name = _ref_is_name(self._columns_field)
-        features = Feature.lookup().dict()
-        for key, field in self._obs_fields.items():
-            feature = features.get(key)
-            registry = field.field.model
-            organism = check_registry_organism(field.field.model, self._organism).get(
-                "organism"
+        return super().validate()
+
+    def to_cellxgene_anndata(
+        self, is_primary_data: bool, title: str | None = None
+    ) -> ad.AnnData:
+        """Converts the AnnData object to the cellxgene-schema input format.
+
+        cellxgene expects the obs fields to be {entity}_ontology_id fields and has many further requirements which are
+        documented here: https://github.com/chanzuckerberg/single-cell-curation/tree/main/schema.
+        This function checks for most but not all requirements of the CELLxGENE schema.
+        If you want to ensure that it fully adheres to the CELLxGENE schema, run `cellxgene-schema` on the AnnData object.
+
+        Args:
+            is_primary_data: Whether the measured data is primary data or not.
+            title: Title of the AnnData object. Commonly the name of the publication.
+
+        Returns:
+            An AnnData object which adheres to the cellxgene-schema.
+        """
+        # Create a copy since we modify the AnnData object extensively
+        adata_cxg = self._adata.copy()
+
+        # cellxgene requires an embedding
+        embedding_pattern = r"^[a-zA-Z][a-zA-Z0-9_.-]*$"
+        exclude_key = "spatial"
+        matching_keys = [
+            key
+            for key in adata_cxg.obsm.keys()
+            if re.match(embedding_pattern, key) and key != exclude_key
+        ]
+        if len(matching_keys) == 0:
+            raise ValueError(
+                "Unable to find an embedding key. Please calculate an embedding."
             )
-            labels = registry.from_values(
-                values=self._validated_values[key], field=field, organism=organism
+
+        # convert name column to ontology_term_id column
+        for column in adata_cxg.obs.columns:
+            if column in self.categoricals and not column.endswith("_ontology_term_id"):
+                mapped_column = self._convert_name_to_ontology_id(
+                    adata_cxg.obs[column], field=self.categoricals.get(column)
+                )
+                if mapped_column is not None:
+                    adata_cxg.obs[f"{column}_ontology_term_id"] = mapped_column
+
+        # drop the name columns for ontologies. cellxgene does not allow them.
+        drop_columns = [
+            i
+            for i in adata_cxg.obs.columns
+            if f"{i}_ontology_term_id" in adata_cxg.obs.columns
+        ]
+        adata_cxg.obs.drop(columns=drop_columns, inplace=True)
+
+        # Add cellxgene metadata to AnnData object
+        if "is_primary_data" not in adata_cxg.obs.columns:
+            adata_cxg.obs["is_primary_data"] = is_primary_data
+        if "feature_is_filtered" not in adata_cxg.var.columns:
+            logger.warn(
+                "column 'feature_is_filtered' not present in var. Setting to default"
+                " value of False."
             )
-            if len(labels) == 0:
-                continue
-            if hasattr(registry, "_name_field"):
-                label_ref_is_name = field.field.name == registry._name_field
-                add_labels(
-                    artifact,
-                    records=labels,
-                    feature=feature,
-                    feature_ref_is_name=feature_ref_is_name,
-                    label_ref_is_name=label_ref_is_name,
-                    from_curator=True,
+            adata_cxg.var["feature_is_filtered"] = False
+        if title is None:
+            raise ValueError("please pass a title!")
+        else:
+            adata_cxg.uns["title"] = title
+        adata_cxg.uns["cxg_lamin_schema_reference"] = self.schema_reference
+        adata_cxg.uns["cxg_lamin_schema_version"] = self.schema_version
+
+        return adata_cxg
+
+
+class ValueUnit:
+    """Base class for handling value-unit combinations."""
+
+    @staticmethod
+    def parse_value_unit(value: str, is_dose: bool = True) -> tuple[str, str] | None:
+        """Parse a string containing a value and unit into a tuple."""
+        if not isinstance(value, str) or not value.strip():
+            return None
+
+        value = str(value).strip()
+        match = re.match(r"^(\d*\.?\d{0,1})\s*([a-zA-ZμµΜ]+)$", value)
+
+        if not match:
+            raise ValueError(
+                f"Invalid format: {value}. Expected format: number with max 1 decimal place + unit"
+            )
+
+        number, unit = match.groups()
+        formatted_number = f"{float(number):.1f}"
+
+        if is_dose:
+            standardized_unit = DoseHandler.standardize_unit(unit)
+            if not DoseHandler.validate_unit(standardized_unit):
+                raise ValueError(
+                    f"Invalid dose unit: {unit}. Must be convertible to one of: nM, μM, mM, M"
+                )
+        else:
+            standardized_unit = TimeHandler.standardize_unit(unit)
+            if not TimeHandler.validate_unit(standardized_unit):
+                raise ValueError(
+                    f"Invalid time unit: {unit}. Must be convertible to one of: h, m, s, d, y"
                 )
 
-        return artifact.save()
+        return formatted_number, standardized_unit
 
 
-class Curator(BaseCurator):
-    """Dataset curator.
+class DoseHandler:
+    """Handler for dose-related operations."""
 
-    A `Curator` object makes it easy to save validated & annotated artifacts.
+    VALID_UNITS = {"nM", "μM", "µM", "mM", "M"}
+    UNIT_MAP = {
+        "nm": "nM",
+        "NM": "nM",
+        "um": "μM",
+        "UM": "μM",
+        "μm": "μM",
+        "μM": "μM",
+        "µm": "μM",
+        "µM": "μM",
+        "mm": "mM",
+        "MM": "mM",
+        "m": "M",
+        "M": "M",
+    }
 
-    Example:
+    @classmethod
+    def validate_unit(cls, unit: str) -> bool:
+        """Validate if the dose unit is acceptable."""
+        return unit in cls.VALID_UNITS
 
-    >>> curator = ln.Curator.from_df(
-    >>>     df,
-    >>>     # define validation criteria as mappings
-    >>>     columns=ln.Feature.name,  # map column names
-    >>>     categoricals={"perturbation": ln.ULabel.name},  # map categories
-    >>> )
-    >>> curator.validate()  # validate the data in df
-    >>> artifact = curator.save_artifact(description="my RNA-seq")
-    >>> artifact.describe()  # see annotations
+    @classmethod
+    def standardize_unit(cls, unit: str) -> str:
+        """Standardize dose unit to standard formats."""
+        return cls.UNIT_MAP.get(unit, unit)
 
-    `curator.validate()` maps values within `df` according to the mapping criteria and logs validated & problematic values.
+    @classmethod
+    def validate_values(cls, values: pd.Series) -> list:
+        """Validate pert_dose values with strict case checking."""
+        errors = []
 
-    If you find non-validated values, you have several options:
+        for idx, value in values.items():
+            if pd.isna(value):
+                continue
 
-    - new values found in the data can be registered using :meth:`~lamindb.core.DataFrameCurator.add_new_from`
-    - non-validated values can be accessed using :meth:`~lamindb.core.DataFrameCurator.non_validated` and addressed manually
-    """
+            if isinstance(value, (int, float)):
+                errors.append(
+                    f"Row {idx} - Missing unit for dose: {value}. Must include a unit (nM, μM, mM, M)"
+                )
+                continue
+
+            try:
+                ValueUnit.parse_value_unit(value, is_dose=True)
+            except ValueError as e:
+                errors.append(f"Row {idx} - {str(e)}")
+
+        return errors
+
+
+class TimeHandler:
+    """Handler for time-related operations."""
+
+    VALID_UNITS = {"h", "m", "s", "d", "y"}
 
     @classmethod
-    @doc_args(DataFrameCurator.__doc__)
-    def from_df(
-        cls,
-        df: pd.DataFrame,
-        categoricals: dict[str, FieldAttr] | None = None,
-        columns: FieldAttr = Feature.name,
-        using_key: str | None = None,
-        verbosity: str = "hint",
-        organism: str | None = None,
-    ) -> DataFrameCurator:
-        """{}"""  # noqa: D415
-        return DataFrameCurator(
-            df=df,
-            categoricals=categoricals,
-            columns=columns,
-            using_key=using_key,
-            verbosity=verbosity,
-            organism=organism,
-        )
+    def validate_unit(cls, unit: str) -> bool:
+        """Validate if the time unit is acceptable."""
+        return unit == unit.lower() and unit in cls.VALID_UNITS
 
     @classmethod
-    @doc_args(AnnDataCurator.__doc__)
-    def from_anndata(
-        cls,
-        data: ad.AnnData | UPathStr,
-        var_index: FieldAttr,
-        categoricals: dict[str, FieldAttr] | None = None,
-        obs_columns: FieldAttr = Feature.name,
-        using_key: str | None = None,
-        verbosity: str = "hint",
-        organism: str | None = None,
-        sources: dict[str, Record] | None = None,
-    ) -> AnnDataCurator:
-        """{}"""  # noqa: D415
-        return AnnDataCurator(
-            data=data,
-            var_index=var_index,
-            categoricals=categoricals,
-            obs_columns=obs_columns,
-            using_key=using_key,
-            verbosity=verbosity,
-            organism=organism,
-            sources=sources,
-        )
+    def standardize_unit(cls, unit: str) -> str:
+        """Standardize time unit to standard formats."""
+        if unit.startswith("hr"):
+            return "h"
+        elif unit.startswith("min"):
+            return "m"
+        elif unit.startswith("sec"):
+            return "s"
+        return unit[0].lower()
 
     @classmethod
-    @doc_args(MuDataCurator.__doc__)
-    def from_mudata(
-        cls,
-        mdata: MuData,
-        var_index: dict[str, dict[str, FieldAttr]],
-        categoricals: dict[str, FieldAttr] | None = None,
-        using_key: str | None = None,
+    def validate_values(cls, values: pd.Series) -> list:
+        """Validate pert_time values."""
+        errors = []
+
+        for idx, value in values.items():
+            if pd.isna(value):
+                continue
+
+            if isinstance(value, (int, float)):
+                errors.append(
+                    f"Row {idx} - Missing unit for time: {value}. Must include a unit (h, m, s, d, y)"
+                )
+                continue
+
+            try:
+                ValueUnit.parse_value_unit(value, is_dose=False)
+            except ValueError as e:
+                errors.append(f"Row {idx} - {str(e)}")
+
+        return errors
+
+
+class PertAnnDataCatManager(CellxGeneAnnDataCatManager):
+    """Curator flow for Perturbation data."""
+
+    PERT_COLUMNS = {"compound", "genetic", "biologic", "physical"}
+
+    def __init__(
+        self,
+        adata: ad.AnnData,
+        organism: Literal["human", "mouse"] = "human",
+        pert_dose: bool = True,
+        pert_time: bool = True,
+        *,
         verbosity: str = "hint",
-        organism: str | None = None,
-    ) -> MuDataCurator:
-        """{}"""  # noqa: D415
-        return MuDataCurator(
-            mdata=mdata,
-            var_index=var_index,
-            categoricals=categoricals,
-            using_key=using_key,
+        cxg_schema_version: Literal["5.0.0", "5.1.0"] = "5.1.0",
+    ):
+        """Initialize the curator with configuration and validation settings."""
+        import bionty as bt
+
+        self._pert_time = pert_time
+        self._pert_dose = pert_dose
+
+        self._validate_initial_data(adata)
+        self._setup_configuration(adata)
+
+        self._setup_sources(adata)
+        self._setup_compound_source()
+
+        super().__init__(
+            adata=adata,
+            categoricals=self.PT_CATEGORICALS,
+            defaults=self.PT_DEFAULT_VALUES,
             verbosity=verbosity,
             organism=organism,
+            extra_sources=self.PT_SOURCES,
+            schema_version=cxg_schema_version,
         )
 
-    @classmethod
-    @doc_args(SOMACurator.__doc__)
-    def from_tiledbsoma(
-        cls,
-        experiment_uri: UPathStr,
-        var_index: dict[str, tuple[str, FieldAttr]],
-        categoricals: dict[str, FieldAttr] | None = None,
-        obs_columns: FieldAttr = Feature.name,
-        using_key: str | None = None,
-        organism: str | None = None,
-        sources: dict[str, Record] | None = None,
-        exclude: dict[str, str | list[str]] | None = None,
-    ) -> SOMACurator:
-        """{}"""  # noqa: D415
-        return SOMACurator(
-            experiment_uri=experiment_uri,
-            var_index=var_index,
-            categoricals=categoricals,
-            obs_columns=obs_columns,
-            using_key=using_key,
-            organism=organism,
-            sources=sources,
-            exclude=exclude,
+    def _setup_configuration(self, adata: ad.AnnData):
+        """Set up default configuration values."""
+        import bionty as bt
+        import wetlab as wl
+
+        self.PT_DEFAULT_VALUES = (
+            CellxGeneAnnDataCatManager._get_categoricals_defaults()
+            | {
+                "cell_line": "unknown",
+                "pert_target": "unknown",
+            }
         )
 
-    @classmethod
-    def from_spatialdata(
-        cls,
-        sdata,
-        var_index: dict[str, FieldAttr],
-        categoricals: dict[str, dict[str, FieldAttr]] | None = None,
-        using_key: str | None = None,
-        organism: str | None = None,
-        sources: dict[str, dict[str, Record]] | None = None,
-        exclude: dict[str, dict] | None = None,
-        verbosity: str = "hint",
-        *,
-        sample_metadata_key: str = "sample",
-    ):
-        """Curation flow for a ``Spatialdata`` object.
+        self.PT_CATEGORICALS = CellxGeneAnnDataCatManager._get_categoricals() | {
+            k: v
+            for k, v in {
+                "cell_line": bt.CellLine.name,
+                "pert_target": wl.PerturbationTarget.name,
+                "pert_genetic": wl.GeneticPerturbation.name,
+                "pert_compound": wl.Compound.name,
+                "pert_biologic": wl.Biologic.name,
+                "pert_physical": wl.EnvironmentalPerturbation.name,
+            }.items()
+            if k in adata.obs.columns
+        }
+        # if "donor_id" in self.PT_CATEGORICALS:
+        #     self.PT_CATEGORICALS["donor_id"] = Donor.name
+
+    def _setup_sources(self, adata: ad.AnnData):
+        """Set up data sources."""
+        self.PT_SOURCES = {}
+        # if "cell_line" in adata.obs.columns:
+        #     self.PT_SOURCES["cell_line"] = (
+        #         bt.Source.filter(name="depmap").first()
+        #     )
+        if "pert_compound" in adata.obs.columns:
+            import bionty as bt
+
+            self.PT_SOURCES["pert_compound"] = bt.Source.filter(
+                entity="wetlab.Compound", name="chebi"
+            ).first()
+
+    def _validate_initial_data(self, adata: ad.AnnData):
+        """Validate the initial data structure."""
+        self._validate_required_columns(adata)
+        self._validate_perturbation_types(adata)
+
+    def _validate_required_columns(self, adata: ad.AnnData):
+        """Validate required columns are present."""
+        if "pert_target" not in adata.obs.columns:
+            if (
+                "pert_name" not in adata.obs.columns
+                or "pert_type" not in adata.obs.columns
+            ):
+                raise ValidationError(
+                    "either 'pert_target' or both 'pert_name' and 'pert_type' must be present"
+                )
+        else:
+            if "pert_name" not in adata.obs.columns:
+                logger.warning(
+                    "no 'pert' column found in adata.obs, will only curate 'pert_target'"
+                )
+            elif "pert_type" not in adata.obs.columns:
+                raise ValidationError("both 'pert' and 'pert_type' must be present")
+
+    def _validate_perturbation_types(self, adata: ad.AnnData):
+        """Validate perturbation types."""
+        if "pert_type" in adata.obs.columns:
+            data_pert_types = set(adata.obs["pert_type"].unique())
+            invalid_pert_types = data_pert_types - self.PERT_COLUMNS
+            if invalid_pert_types:
+                raise ValidationError(
+                    f"invalid pert_type found: {invalid_pert_types}!\n"
+                    f"    → allowed values: {self.PERT_COLUMNS}"
+                )
+            self._process_perturbation_types(adata, data_pert_types)
+
+    def _process_perturbation_types(self, adata: ad.AnnData, pert_types: set):
+        """Process and map perturbation types."""
+        for pert_type in pert_types:
+            col_name = "pert_" + pert_type
+            adata.obs[col_name] = adata.obs["pert_name"].where(
+                adata.obs["pert_type"] == pert_type, None
+            )
+            if adata.obs[col_name].dtype.name == "category":
+                adata.obs[col_name].cat.remove_unused_categories()
+            logger.important(f"mapped 'pert_name' to '{col_name}'")
 
-        See also :class:`~lamindb.Curator`.
+    def _setup_compound_source(self):
+        """Set up the compound source with muted logging."""
+        import bionty as bt
+        import wetlab as wl
+
+        with logger.mute():
+            chebi_source = bt.Source.filter(
+                entity="wetlab.Compound", name="chebi"
+            ).first()
+            if not chebi_source:
+                wl.Compound.add_source(
+                    bt.Source.filter(entity="Drug", name="chebi").first()
+                )
 
-        Note that if genes or other measurements are removed from the SpatialData object,
-        the object should be recreated.
+    def validate(self) -> bool:  # type: ignore
+        """Validate the AnnData object."""
+        validated = super().validate()
+
+        if self._pert_dose:
+            validated &= self._validate_dose_column()
+        if self._pert_time:
+            validated &= self._validate_time_column()
+
+        self._is_validated = validated
+
+        # sort columns
+        first_columns = [
+            "pert_target",
+            "pert_genetic",
+            "pert_compound",
+            "pert_biologic",
+            "pert_physical",
+            "pert_dose",
+            "pert_time",
+            "organism",
+            "cell_line",
+            "cell_type",
+            "disease",
+            "tissue_type",
+            "tissue",
+            "assay",
+            "suspension_type",
+            "donor_id",
+            "sex",
+            "self_reported_ethnicity",
+            "development_stage",
+            "pert_name",
+            "pert_type",
+        ]
+        sorted_columns = [
+            col for col in first_columns if col in self._adata.obs.columns
+        ] + [col for col in self._adata.obs.columns if col not in first_columns]
+        # must assign to self._df to ensure .standardize works correctly
+        self._obs_df = self._adata.obs[sorted_columns]
+        self._adata.obs = self._obs_df
+        return validated
+
+    def standardize(self, key: str) -> pd.DataFrame:
+        """Standardize the AnnData object."""
+        super().standardize(key)
+        self._adata.obs = self._obs_df
+
+    def _validate_dose_column(self) -> bool:
+        """Validate the dose column."""
+        if not Feature.filter(name="pert_dose").exists():
+            Feature(name="pert_dose", dtype="str").save()  # type: ignore
+
+        dose_errors = DoseHandler.validate_values(self._adata.obs["pert_dose"])
+        if dose_errors:
+            self._log_validation_errors("pert_dose", dose_errors)
+            return False
+        return True
+
+    def _validate_time_column(self) -> bool:
+        """Validate the time column."""
+        if not Feature.filter(name="pert_time").exists():
+            Feature(name="pert_time", dtype="str").save()  # type: ignore
+
+        time_errors = TimeHandler.validate_values(self._adata.obs["pert_time"])
+        if time_errors:
+            self._log_validation_errors("pert_time", time_errors)
+            return False
+        return True
+
+    def _log_validation_errors(self, column: str, errors: list):
+        """Log validation errors with formatting."""
+        errors_print = "\n    ".join(errors)
+        logger.warning(
+            f"invalid {column} values found!\n    {errors_print}\n"
+            f"    → run {colors.cyan('standardize_dose_time()')}"
+        )
 
-        In the following docstring, an accessor refers to either a ``.table`` key or the ``sample_metadata_key``.
+    def standardize_dose_time(self) -> pd.DataFrame:
+        """Standardize dose and time values."""
+        standardized_df = self._adata.obs.copy()
 
-        Args:
-            sdata: The SpatialData object to curate.
-            var_index: A dictionary mapping table keys to the ``.var`` indices.
-            categoricals: A nested dictionary mapping an accessor to dictionaries that map columns to a registry field.
-            using_key: A reference LaminDB instance.
-            organism: The organism name.
-            sources: A dictionary mapping an accessor to dictionaries that map columns to Source records.
-            exclude: A dictionary mapping an accessor to dictionaries of column names to values to exclude from validation.
-                When specific :class:`~bionty.Source` instances are pinned and may lack default values (e.g., "unknown" or "na"),
-                using the exclude parameter ensures they are not validated.
-            verbosity: The verbosity level of the logger.
-            sample_metadata_key: The key in ``.attrs`` that stores the sample level metadata.
-
-        Examples:
-            >>> import lamindb as ln
-            >>> import bionty as bt
-            >>> curator = ln.Curator.from_spatialdata(
-            ...     sdata,
-            ...     var_index={
-            ...         "table_1": bt.Gene.ensembl_gene_id,
-            ...     },
-            ...     categoricals={
-            ...         "table1":
-            ...             {"cell_type_ontology_id": bt.CellType.ontology_id, "donor_id": ln.ULabel.name},
-            ...         "sample":
-            ...             {"experimental_factor": bt.ExperimentalFactor.name},
-            ...     },
-            ...     organism="human",
-            ... )
-        """
-        try:
-            import spatialdata
-        except ImportError as e:
-            raise ImportError(
-                "Please install spatialdata: pip install spatialdata"
-            ) from e
+        if "pert_dose" in self._adata.obs.columns:
+            standardized_df = self._standardize_column(
+                standardized_df, "pert_dose", is_dose=True
+            )
 
-        from ._spatial import SpatialDataCurator
+        if "pert_time" in self._adata.obs.columns:
+            standardized_df = self._standardize_column(
+                standardized_df, "pert_time", is_dose=False
+            )
 
-        return SpatialDataCurator(
-            sdata=sdata,
-            var_index=var_index,
-            categoricals=categoricals,
-            using_key=using_key,
-            verbosity=verbosity,
-            organism=organism,
-            sources=sources,
-            exclude=exclude,
-            sample_metadata_key=sample_metadata_key,
-        )
+        self._adata.obs = standardized_df
+        return standardized_df
+
+    def _standardize_column(
+        self, df: pd.DataFrame, column: str, is_dose: bool
+    ) -> pd.DataFrame:
+        """Standardize values in a specific column."""
+        for idx, value in self._adata.obs[column].items():
+            if pd.isna(value) or (
+                isinstance(value, str) and (not value.strip() or value.lower() == "nan")
+            ):
+                df.at[idx, column] = None
+                continue
 
+            try:
+                num, unit = ValueUnit.parse_value_unit(value, is_dose=is_dose)
+                df.at[idx, column] = f"{num}{unit}"
+            except ValueError:
+                continue
 
-def get_registry_instance(registry: Record, using_key: str | None = None) -> Record:
-    """Get a registry instance using a specific instance."""
-    if using_key is not None and using_key != "default":
-        return registry.using(using_key)
-    return registry
+        return df
 
 
 def get_current_filter_kwargs(registry: type[Record], kwargs: dict) -> dict:
@@ -1871,11 +3180,11 @@ def validate_categories(
     values: Iterable[str],
     field: FieldAttr,
     key: str,
-    using_key: str | None = None,
     organism: str | None = None,
     source: Record | None = None,
     exclude: str | list | None = None,
     hint_print: str | None = None,
+    curator: CatManager | None = None,
 ) -> tuple[bool, list]:
     """Validate ontology terms in a pandas series using LaminDB registries.
 
@@ -1883,7 +3192,6 @@ def validate_categories(
         values: The values to validate.
         field: The field attribute.
         key: The key referencing the slot in the DataFrame.
-        using_key: A reference LaminDB instance.
         organism: The organism name.
         source: The source record.
         exclude: Exclude specific values from validation.
@@ -1918,22 +3226,8 @@ def validate_categories(
     non_validated = inspect_result.non_validated
     syn_mapper = inspect_result.synonyms_mapper
 
-    # inspect the non-validated values from the using_key instance
-    values_validated = []
-    if using_key is not None and using_key != "default" and non_validated:
-        registry_using = get_registry_instance(registry, using_key)
-        inspect_result = inspect_instance(
-            values=non_validated,
-            field=field,
-            registry=registry_using,
-            exclude=exclude,
-            **kwargs,
-        )
-        non_validated = inspect_result.non_validated
-        values_validated += inspect_result.validated
-        syn_mapper.update(inspect_result.synonyms_mapper)
-
     # inspect the non-validated values from public (bionty only)
+    values_validated = []
     if hasattr(registry, "public"):
         verbosity = settings.verbosity
         try:
@@ -1975,6 +3269,10 @@ def validate_categories(
         if logger.indent == "":
             _log_mapping_info()
         logger.warning(warning_message)
+        if curator is not None:
+            curator._validate_category_error_messages = strip_ansi_codes(
+                warning_message
+            )
         logger.indent = ""
         return False, non_validated
 
@@ -1982,7 +3280,6 @@ def validate_categories(
 def standardize_categories(
     values: Iterable[str],
     field: FieldAttr,
-    using_key: str | None = None,
     organism: str | None = None,
     source: Record | None = None,
 ) -> dict:
@@ -1999,30 +3296,15 @@ def standardize_categories(
         mute=True,
         return_mapper=True,
     )
-
-    if len(values) > len(syn_mapper):  # type: ignore
-        # standardize values using the using_key instance
-        if using_key is not None and using_key != "default":
-            registry_using = get_registry_instance(registry, using_key)
-            syn_mapper.update(
-                registry_using.standardize(
-                    [v for v in values if v not in syn_mapper],
-                    field=field.field.name,
-                    organism=organism,
-                    source=source,
-                    mute=True,
-                    return_mapper=True,
-                )
-            )
     return syn_mapper
 
 
 def validate_categories_in_df(
     df: pd.DataFrame,
     fields: dict[str, FieldAttr],
-    using_key: str | None = None,
     sources: dict[str, Record] = None,
     exclude: dict | None = None,
+    curator: CatManager | None = None,
     **kwargs,
 ) -> tuple[bool, dict]:
     """Validate categories in DataFrame columns using LaminDB registries."""
@@ -2038,9 +3320,9 @@ def validate_categories_in_df(
             df[key],
             field=field,
             key=key,
-            using_key=using_key,
             source=sources.get(key),
             exclude=exclude.get(key) if exclude else None,
+            curator=curator,
             **kwargs,
         )
         validated &= is_val
@@ -2055,80 +3337,72 @@ def save_artifact(
     columns_field: FieldAttr | dict[str, FieldAttr],
     description: str | None = None,
     organism: str | None = None,
-    adata: ad.AnnData | None = None,
     key: str | None = None,
+    artifact: Artifact | None = None,
     revises: Artifact | None = None,
     run: Run | None = None,
+    schema: Schema | None = None,
 ) -> Artifact:
     """Save all metadata with an Artifact.
 
     Args:
-        data: The DataFrame or AnnData object to save.
+        data: The DataFrame/AnnData/MuData object to save.
         fields: A dictionary mapping obs_column to registry_field.
         columns_field: The registry field to validate variables index against.
         description: A description of the artifact.
         organism: The organism name.
-        adata: The AnnData object to save and get n_observations, must be provided if data is a path.
         type: The artifact type.
-        key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`. Artifacts with the same key form a revision family.
+        key: A path-like key to reference artifact in default storage, e.g., `"myfolder/myfile.fcs"`. Artifacts with the same key form a version family.
+        artifact: A already registered artifact. Passing this will not save a new artifact from data.
         revises: Previous version of the artifact. Triggers a revision.
         run: The run that creates the artifact.
 
     Returns:
         The saved Artifact.
     """
-    from .._artifact import data_is_anndata
+    from .._artifact import data_is_anndata, data_is_mudata
     from ..core._data import add_labels
 
-    artifact = None
-    if data_is_anndata(data):
-        assert adata is not None  # noqa: S101
-        artifact = Artifact.from_anndata(
-            data, description=description, key=key, revises=revises, run=run
-        )
-        artifact.n_observations = adata.shape[0]
-        data = adata
-
-    elif isinstance(data, pd.DataFrame):
-        artifact = Artifact.from_df(
-            data, description=description, key=key, revises=revises, run=run
-        )
-    else:
-        try:
-            from mudata import MuData
-
-            if isinstance(data, MuData):
-                artifact = Artifact.from_mudata(
-                    data,
-                    description=description,
-                    key=key,
-                    revises=revises,
-                    run=run,
-                )
-                artifact.n_observations = data.n_obs
-        except ImportError:
-            pass
     if artifact is None:
-        raise ValueError("data must be a DataFrame, AnnData or MuData object.")
+        if data_is_anndata(data):
+            artifact = Artifact.from_anndata(
+                data, description=description, key=key, revises=revises, run=run
+            )
+        elif isinstance(data, pd.DataFrame):
+            artifact = Artifact.from_df(
+                data, description=description, key=key, revises=revises, run=run
+            )
+        elif data_is_mudata(data):
+            artifact = Artifact.from_mudata(
+                data,
+                description=description,
+                key=key,
+                revises=revises,
+                run=run,
+            )
+    artifact.schema = schema
     artifact.save()
 
-    feature_kwargs = check_registry_organism(
-        (
-            list(columns_field.values())[0].field.model
-            if isinstance(columns_field, dict)
-            else columns_field.field.model
-        ),
-        organism,
-    )
+    if organism is not None:
+        feature_kwargs = check_registry_organism(
+            (
+                list(columns_field.values())[0].field.model
+                if isinstance(columns_field, dict)
+                else columns_field.field.model
+            ),
+            organism,
+        )
+    else:
+        feature_kwargs = {}
 
     if artifact.otype == "DataFrame":
-        artifact.features._add_set_from_df(field=columns_field, **feature_kwargs)
+        artifact.features._add_set_from_df(field=columns_field, **feature_kwargs)  # type: ignore
     elif artifact.otype == "AnnData":
-        artifact.features._add_set_from_anndata(
+        artifact.features._add_set_from_anndata(  # type: ignore
             var_field=columns_field, **feature_kwargs
         )
     elif artifact.otype == "MuData":
-        artifact.features._add_set_from_mudata(
+        artifact.features._add_set_from_mudata(  # type: ignore
             var_fields=columns_field, **feature_kwargs
         )
     else:
@@ -2202,7 +3476,7 @@ def save_artifact(
         )
 
     slug = ln_setup.settings.instance.slug
-    if ln_setup.settings.instance.is_remote:  # pragma: no cover
+    if ln_setup.settings.instance.is_remote:  # pdagma: no cover
         logger.important(f"go to https://lamin.ai/{slug}/artifact/{artifact.uid}")
     return artifact
 
@@ -2224,7 +3498,6 @@ def update_registry(
     values: list[str],
     field: FieldAttr,
     key: str,
-    using_key: str | None = None,
     validated_only: bool = True,
     df: pd.DataFrame | None = None,
     organism: str | None = None,
@@ -2233,13 +3506,12 @@ def update_registry(
     exclude: str | list | None = None,
     **kwargs,
 ) -> None:
-    """Save features or labels records in the default instance from the using_key instance.
+    """Save features or labels records in the default instance..
 
     Args:
         values: A list of values to be saved as labels.
         field: The FieldAttr object representing the field for which labels are being saved.
         key: The name of the feature to save.
-        using_key: The name of the instance from which to transfer labels (if applicable).
         validated_only: If True, only save validated labels.
         df: A DataFrame to save labels from.
         organism: The organism name.
@@ -2290,22 +3562,10 @@ def update_registry(
             i for i in values if i not in existing_and_public_labels
         ]
 
-        # inspect and save validated records the using_key instance
-        (
-            labels_saved[f"from {using_key}"],
-            non_validated_labels,
-        ) = update_registry_from_using_instance(
-            non_validated_labels,
-            field=field,
-            using_key=using_key,
-            exclude=exclude,
-            **filter_kwargs,
-        )
-
         # save non-validated/new records
         labels_saved["new"] = non_validated_labels
         if not validated_only:
-            non_validated_records = []
+            non_validated_records: RecordList[Any] = []  # type: ignore
             if df is not None and registry == Feature:
                 nonval_columns = Feature.inspect(df.columns, mute=True).non_validated
                 non_validated_records = Feature.from_df(df.loc[:, nonval_columns])
@@ -2379,48 +3639,6 @@ def save_ulabels_parent(values: list[str], field: FieldAttr, key: str) -> None:
     is_feature.children.add(*all_records)
 
 
-def update_registry_from_using_instance(
-    values: list[str],
-    field: FieldAttr,
-    using_key: str | None = None,
-    exclude: str | list | None = None,
-    **kwargs,
-) -> tuple[list[str], list[str]]:
-    """Save features or labels records from the using_key instance.
-
-    Args:
-        values: A list of values to be saved as labels.
-        field: The FieldAttr object representing the field for which labels are being saved.
-        using_key: The name of the instance from which to transfer labels (if applicable).
-        kwargs: Additional keyword arguments to pass to the registry model.
-
-    Returns:
-        A tuple containing the list of saved labels and the list of non-saved labels.
-    """
-    labels_saved = []
-    not_saved = values
-
-    if using_key is not None and using_key != "default":
-        registry_using = get_registry_instance(field.field.model, using_key)
-
-        inspect_result_using = inspect_instance(
-            values=values,
-            field=field,
-            registry=registry_using,
-            exclude=exclude,
-            **kwargs,
-        )
-        labels_using = registry_using.filter(
-            **{f"{field.field.name}__in": inspect_result_using.validated}
-        ).all()
-        for label_using in labels_using:
-            label_using.save()
-            labels_saved.append(getattr(label_using, field.field.name))
-        not_saved = inspect_result_using.non_validated
-
-    return labels_saved, not_saved
-
-
 def _save_organism(name: str):
     """Save an organism record."""
     import bionty as bt
@@ -2445,4 +3663,121 @@ def _ref_is_name(field: FieldAttr) -> bool | None:
     return field.field.name == name_field
 
 
-Curate = Curator  # backward compat
+# backward compat constructors ------------------
+
+
+@classmethod  # type: ignore
+def from_df(
+    cls,
+    df: pd.DataFrame,
+    categoricals: dict[str, FieldAttr] | None = None,
+    columns: FieldAttr = Feature.name,
+    verbosity: str = "hint",
+    organism: str | None = None,
+) -> DataFrameCatManager:
+    return DataFrameCatManager(
+        df=df,
+        categoricals=categoricals,
+        columns=columns,
+        verbosity=verbosity,
+        organism=organism,
+    )
+
+
+@classmethod  # type: ignore
+def from_anndata(
+    cls,
+    data: ad.AnnData | UPathStr,
+    var_index: FieldAttr,
+    categoricals: dict[str, FieldAttr] | None = None,
+    obs_columns: FieldAttr = Feature.name,
+    verbosity: str = "hint",
+    organism: str | None = None,
+    sources: dict[str, Record] | None = None,
+) -> AnnDataCatManager:
+    return AnnDataCatManager(
+        data=data,
+        var_index=var_index,
+        categoricals=categoricals,
+        obs_columns=obs_columns,
+        verbosity=verbosity,
+        organism=organism,
+        sources=sources,
+    )
+
+
+@classmethod  # type: ignore
+def from_mudata(
+    cls,
+    mdata: MuData,
+    var_index: dict[str, dict[str, FieldAttr]],
+    categoricals: dict[str, FieldAttr] | None = None,
+    verbosity: str = "hint",
+    organism: str | None = None,
+) -> MuDataCatManager:
+    return MuDataCatManager(
+        mdata=mdata,
+        var_index=var_index,
+        categoricals=categoricals,
+        verbosity=verbosity,
+        organism=organism,
+    )
+
+
+@classmethod  # type: ignore
+def from_tiledbsoma(
+    cls,
+    experiment_uri: UPathStr,
+    var_index: dict[str, tuple[str, FieldAttr]],
+    categoricals: dict[str, FieldAttr] | None = None,
+    obs_columns: FieldAttr = Feature.name,
+    organism: str | None = None,
+    sources: dict[str, Record] | None = None,
+    exclude: dict[str, str | list[str]] | None = None,
+) -> TiledbsomaCatManager:
+    return TiledbsomaCatManager(
+        experiment_uri=experiment_uri,
+        var_index=var_index,
+        categoricals=categoricals,
+        obs_columns=obs_columns,
+        organism=organism,
+        sources=sources,
+        exclude=exclude,
+    )
+
+
+@classmethod  # type: ignore
+def from_spatialdata(
+    cls,
+    sdata,
+    var_index: dict[str, FieldAttr],
+    categoricals: dict[str, dict[str, FieldAttr]] | None = None,
+    organism: str | None = None,
+    sources: dict[str, dict[str, Record]] | None = None,
+    exclude: dict[str, dict] | None = None,
+    verbosity: str = "hint",
+    *,
+    sample_metadata_key: str = "sample",
+):
+    try:
+        import spatialdata
+    except ImportError as e:
+        raise ImportError("Please install spatialdata: pip install spatialdata") from e
+
+    return SpatialDataCatManager(
+        sdata=sdata,
+        var_index=var_index,
+        categoricals=categoricals,
+        verbosity=verbosity,
+        organism=organism,
+        sources=sources,
+        exclude=exclude,
+        sample_metadata_key=sample_metadata_key,
+    )
+
+
+CatManager.from_df = from_df  # type: ignore
+CatManager.from_anndata = from_anndata  # type: ignore
+CatManager.from_mudata = from_mudata  # type: ignore
+CatManager.from_spatialdata = from_spatialdata  # type: ignore
+CatManager.from_tiledbsoma = from_tiledbsoma  # type: ignore