martinkim0 commited on
Commit
f5d066e
1 Parent(s): d1d7698

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +98 -0
README.md ADDED
@@ -0,0 +1,98 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: cc-by-4.0
3
+ library_name: scvi-tools
4
+ tags:
5
+ - biology
6
+ - genomics
7
+ - single-cell
8
+ - model_cls_name:SCVI
9
+ - scvi_version:0.20.0
10
+ - anndata_version:0.8.0
11
+ - modality:rna
12
+ - tissue:Bladder
13
+ - annotated:True
14
+ ---
15
+
16
+ # Description
17
+
18
+ Tabula sapiens. An across organ dataset of cell-types in human tissues.
19
+
20
+ # Model properties
21
+
22
+ Many model properties are in the model tags. Some more are listed below.
23
+
24
+ **model_init_params**:
25
+ ```json
26
+ {
27
+ "n_hidden": 128,
28
+ "n_latent": 20,
29
+ "n_layers": 3,
30
+ "dropout_rate": 0.05,
31
+ "dispersion": "gene",
32
+ "gene_likelihood": "nb",
33
+ "latent_distribution": "normal",
34
+ "use_batch_norm": "none",
35
+ "use_layer_norm": "both",
36
+ "encode_covariates": true
37
+ }
38
+ ```
39
+
40
+ **model_setup_anndata_args**:
41
+ ```json
42
+ {
43
+ "layer": null,
44
+ "batch_key": "donor_assay",
45
+ "labels_key": "cell_ontology_class",
46
+ "size_factor_key": null,
47
+ "categorical_covariate_keys": null,
48
+ "continuous_covariate_keys": null
49
+ }
50
+ ```
51
+
52
+ **model_summary_stats**:
53
+ | Summary Stat Key | Value |
54
+ |--------------------------|-------|
55
+ | n_batch | 5 |
56
+ | n_cells | 24583 |
57
+ | n_extra_categorical_covs | 0 |
58
+ | n_extra_continuous_covs | 0 |
59
+ | n_labels | 15 |
60
+ | n_latent_qzm | 20 |
61
+ | n_latent_qzv | 20 |
62
+ | n_vars | 4000 |
63
+
64
+ **model_data_registry**:
65
+ | Registry Key | scvi-tools Location |
66
+ |-------------------|--------------------------------------|
67
+ | X | adata.X |
68
+ | batch | adata.obs['_scvi_batch'] |
69
+ | labels | adata.obs['_scvi_labels'] |
70
+ | latent_qzm | adata.obsm['_scvi_latent_qzm'] |
71
+ | latent_qzv | adata.obsm['_scvi_latent_qzv'] |
72
+ | minify_type | adata.uns['_scvi_adata_minify_type'] |
73
+ | observed_lib_size | adata.obs['_scvi_observed_lib_size'] |
74
+
75
+ **model_parent_module**: scvi.model
76
+
77
+ **data_is_minified**: True
78
+
79
+ # Training data
80
+
81
+ This is an optional link to where the training data is stored if it is too large
82
+ to host on the huggingface Model hub.
83
+
84
+ <!-- If your model is not uploaded with any data (e.g., minified data) on the Model Hub, then make
85
+ sure to provide this field if you want users to be able to access your training data. See the scvi-tools
86
+ documentation for details. -->
87
+
88
+ Training data url: https://zenodo.org/api/files/c676cd14-97a7-4ea6-ba70-2c7faa1d61d8/Bladder_training_data.h5ad
89
+
90
+ # Training code
91
+
92
+ This is an optional link to the code used to train the model.
93
+
94
+ Training code url: https://github.com/scvi-hub-references/tabula_sapiens/main.py
95
+
96
+ # References
97
+
98
+ The Tabula Sapiens: A multi-organ, single-cell transcriptomic atlas of humans. The Tabula Sapiens Consortium. Science 2022.05.13; doi: https: //doi.org/10.1126/science.abl4896