Spatial¶

Note

Please see:

This page is for now just a stub.

# !pip install 'lamindb[jupyter,bionty]'
!lamin init --storage ./test-spatial --schema bionty

import lamindb as ln
import bionty as bt
import matplotlib.pyplot as plt
import scanpy as sc

ln.settings.transform.stem_uid = "daeFs3PkquDW"
ln.settings.transform.version = "draft"
ln.track()

💡 notebook imports: bionty==0.47.1 lamindb==0.75.0 matplotlib==3.9.0 scanpy==1.9.6

💡 saved: Transform(uid='daeFs3PkquDW7a3V', version='draft', name='Spatial', key='spatial', type='notebook', created_by_id=1, updated_at='2024-08-05 13:27:50 UTC')

💡 saved: Run(uid='uBzalHUVG0kMvaXxGKfX', transform_id=1, created_by_id=1)

Run(uid='uBzalHUVG0kMvaXxGKfX', started_at='2024-08-05 13:27:50 UTC', is_consecutive=True, transform_id=1, created_by_id=1)

An example spatial dataset¶

Here, we have a spatial gene expression dataset measured using Visium from Suo22.

This collection contains two parts:

a high-res image of a slice of fetal liver
a single cell expression dataset in .h5ad

img_path = ln.core.datasets.file_tiff_suo22()
img = plt.imread(img_path)
plt.imshow(img)
plt.show()

_images/42e19325a0bdf0a25162edac62d522b7f4aeeb387ec4747fa70d903d17c6d257.png

adata = ln.core.datasets.anndata_suo22_Visium10X()
# subset to the same image
adata = adata[adata.obs["img_id"] == "F121_LP1_4LIV"].copy()
adata

AnnData object with n_obs × n_vars = 3027 × 191
    obs: 'in_tissue', 'array_row', 'array_col', 'sample', 'n_genes_by_counts', 'log1p_n_genes_by_counts', 'total_counts', 'log1p_total_counts', 'pct_counts_in_top_50_genes', 'pct_counts_in_top_100_genes', 'pct_counts_in_top_200_genes', 'pct_counts_in_top_500_genes', 'mt_frac', 'img_id', 'EXP_id', 'Organ', 'Fetal_id', 'SN', 'Visium_Area_id', 'Age_PCW', 'Digestion time', 'paths', 'sample_id', '_scvi_batch', '_scvi_labels', '_indices', 'total_cell_abundance'
    var: 'feature_types', 'genome', 'SYMBOL', 'mt'
    obsm: 'NMF', 'means_cell_abundance_w_sf', 'q05_cell_abundance_w_sf', 'q95_cell_abundance_w_sf', 'spatial', 'stds_cell_abundance_w_sf'

# plot where CD45+ leukocytes are in the slice
sc.pl.scatter(adata, "array_row", "array_col", color="ENSG00000081237")

_images/179081c565bc82cbb6d6b2d06d29e8f57d7ab0ea73263185e4985ff5e28aa665.png

Validate annotations¶

We’ll register the single-cell data and the image as a Collection.

curate = ln.Curate.from_anndata(adata, var_index=bt.Gene.ensembl_gene_id, categoricals={"sample": ln.ULabel.name}, organism="human")

curate.validate()

✅ var_index is validated against Gene.ensembl_gene_id

💡 mapping sample on ULabel.name

❗    1 terms is not validated: 'WSSS_F_IMMsp9838712'
      → save terms via .add_new_from('sample')

False

curate.add_new_from('sample')

✅ added 1 record with ULabel.name for sample: 'WSSS_F_IMMsp9838712'

curate.validate()

✅ var_index is validated against Gene.ensembl_gene_id

✅ sample is validated against ULabel.name

True

Register curated artifact¶

artifact_ad = curate.save_artifact(description="Suo22 Visium10X image F121_LP1_4LIV")

artifact_ad.describe()

Artifact(uid='FIpG2yB33beie1vaNmON', description='Suo22 Visium10X image F121_LP1_4LIV', suffix='.h5ad', type='dataset', _accessor='AnnData', size=9743793, hash='MRyvckic_gbrV_hHpHOlAQ', _hash_type='md5', n_observations=3027, visibility=1, _key_is_virtual=True, updated_at='2024-08-05 13:28:10 UTC')
  Provenance
    .created_by = 'testuser1'
    .storage = '/home/runner/work/lamin-usecases/lamin-usecases/docs/test-spatial'
    .transform = 'Spatial'
    .run = '2024-08-05 13:27:50 UTC'
  Labels
    .ulabels = 'WSSS_F_IMMsp9838712'
  Features
    'sample' = 'WSSS_F_IMMsp9838712'
  Feature sets
    'var' = 'CD99', 'CD38', 'CDC27', 'CD79B', 'CDKL5', 'CDK11A', 'CD9', 'CD4', 'CD22', 'CD6', 'CD74', 'CD44', 'CDH1', 'CDK17', 'CDH3', 'CDK13', 'CD84', 'CDC42', 'CDH19', 'CD5L'
    'obs' = 'sample'

Register a collection¶

artifact_img = ln.Artifact(img_path, description="Suo22 image F121_LP1_4LIV")
artifact_img.save()

collection = ln.Collection([artifact_ad, artifact_img], name="Suo22")
collection.save()

Collection(uid='bXl2owreXlfov4mslR1Q', name='Suo22', hash='8BBgP1aPBOFv9jCebMhGdw', visibility=1, created_by_id=1, transform_id=1, run_id=1, updated_at='2024-08-05 13:28:10 UTC')

# clean up test instance
!lamin delete --force test-spatial
!rm -r test-flow