`sc.external.pp.scrublet` Key error due to internal sc.pp.filter_cells/genes

• I have checked that this issue has not already been reported.
• I have confirmed this bug exists on the latest version of scanpy.
• (optional) I have confirmed this bug exists on the master branch of scanpy.

Note: Please read this guide detailing how to provide the necessary information for us to reproduce your bug.

_run_scrublet sometimes removes cells, due to these lines. As a result, this line results in a key error since the scrubbed data contains fewer cells than the original adata.

Minimal code sample (that we can copy&paste without having any data)

I do not have a code sample that does not require data, but any anndata that contains cells with fewer than 3 genes should trigger the error.

import scanpy
adata = scanpy.read("test.h5ad") # e.g. contains cells with fewer than 3 genes
sc.external.pp.scrublet(test_adata, batch_key = "label")

~/.local/lib/python3.9/site-packages/scanpy/preprocessing/_simple.py:251: ImplicitModificationWarning: Trying to modify attribute `.var` of view, initializing view as actual.
  adata.var['n_cells'] = number
~/.local/lib/python3.9/site-packages/scanpy/preprocessing/_normalization.py:170: UserWarning: Received a view of an AnnData. Making a copy.
  view_to_actual(adata)

Automatically set threshold at doublet score = 0.16
Detected doublet rate = 6.4%
Estimated detectable doublet fraction = 61.7%
Overall doublet rate:
	Expected   = 5.0%
	Estimated  = 10.4%

~/.local/lib/python3.9/site-packages/scanpy/preprocessing/_simple.py:251: ImplicitModificationWarning: Trying to modify attribute `.var` of view, initializing view as actual.
  adata.var['n_cells'] = number
~/.local/lib/python3.9/site-packages/scanpy/preprocessing/_normalization.py:170: UserWarning: Received a view of an AnnData. Making a copy.
  view_to_actual(adata)

Automatically set threshold at doublet score = 0.17
Detected doublet rate = 5.8%
Estimated detectable doublet fraction = 55.7%
Overall doublet rate:
	Expected   = 5.0%
	Estimated  = 10.5%

---------------------------------------------------------------------------
KeyError                                  Traceback (most recent call last)
/tmp/ipykernel_2434661/1304834185.py in <module>
----> 1 sc.external.pp.scrublet(test_adata, batch_key = "label")

~/.local/lib/python3.9/site-packages/scanpy/external/pp/_scrublet.py in scrublet(adata, adata_sim, batch_key, sim_doublet_ratio, expected_doublet_rate, stdev_doublet_rate, synthetic_doublet_umi_subsampling, knn_dist_metric, normalize_variance, log_transform, mean_center, n_prin_comps, use_approx_neighbors, get_doublet_neighbor_parents, n_neighbors, threshold, verbose, copy, random_state)
    250         # Now reset the obs to get the scrublet scores
    251 
--> 252         adata.obs = scrubbed_obs.loc[adata.obs_names.values]
    253 
    254         # Save the .uns from each batch separately

<path>/lib/python3.9/site-packages/pandas/core/indexing.py in __getitem__(self, key)
    929 
    930             maybe_callable = com.apply_if_callable(key, self.obj)
--> 931             return self._getitem_axis(maybe_callable, axis=axis)
    932 
    933     def _is_scalar_access(self, key: tuple):

<path>/lib/python3.9/site-packages/pandas/core/indexing.py in _getitem_axis(self, key, axis)
   1151                     raise ValueError("Cannot index with multidimensional key")
   1152 
-> 1153                 return self._getitem_iterable(key, axis=axis)
   1154 
   1155             # nested tuple slicing

<path>/lib/python3.9/site-packages/pandas/core/indexing.py in _getitem_iterable(self, key, axis)
   1091 
   1092         # A collection of keys
-> 1093         keyarr, indexer = self._get_listlike_indexer(key, axis)
   1094         return self.obj._reindex_with_indexers(
   1095             {axis: [keyarr, indexer]}, copy=True, allow_dups=True

<path>/lib/python3.9/site-packages/pandas/core/indexing.py in _get_listlike_indexer(self, key, axis)
   1312             keyarr, indexer, new_indexer = ax._reindex_non_unique(keyarr)
   1313 
-> 1314         self._validate_read_indexer(keyarr, indexer, axis)
   1315 
   1316         if needs_i8_conversion(ax.dtype) or isinstance(

<path>/lib/python3.9/site-packages/pandas/core/indexing.py in _validate_read_indexer(self, key, indexer, axis)
   1375 
   1376             not_found = list(ensure_index(key)[missing_mask.nonzero()[0]].unique())
-> 1377             raise KeyError(f"{not_found} not in index")
   1378 
   1379 

KeyError: "['GGAACCCTCTCCCAGC-batch1'] not in index"

Versions

Start in scanpy/external/pp/_scrublet.py, especially lines 175-176 where filtering occurs and line 252 where scrubbed_obs is indexed. Reproduce with an AnnData object containing cells with fewer than 3 genes and batch_key="label"; done means sc.external.pp.scrublet completes without a KeyError and preserves consistent observation data.