File size: 14,963 Bytes
8454617
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a90e0da
 
 
 
95d2d5a
 
a90e0da
95d2d5a
a90e0da
 
 
 
 
 
 
 
 
 
 
8454617
 
 
 
 
 
a90e0da
 
 
 
95d2d5a
a90e0da
 
 
 
 
 
 
 
95d2d5a
a90e0da
 
95d2d5a
a90e0da
 
8454617
 
 
 
 
 
 
 
 
 
 
 
 
 
a90e0da
 
 
 
95d2d5a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a90e0da
 
 
 
 
 
95d2d5a
a90e0da
95d2d5a
 
a90e0da
95d2d5a
a90e0da
 
 
95d2d5a
 
a90e0da
95d2d5a
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a90e0da
 
8454617
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
---
library_name: scvi-tools
license: cc-by-4.0
tags:
- biology
- genomics
- single-cell
- model_cls_name:TOTALVI
- scvi_version:1.2.0
- anndata_version:0.11.1
- modality:rna
- modality:protein
- tissue:thymus
- annotated:True
---


TotalVI is a variational inference model for single-cell RNA-seq as well as protein data that can
learn an underlying latent space, integrate technical batches, impute dropouts,
and predict protein expression given gene expression or missing protein data given gene expression
and protein data for a subset of proteins.
The learned low-dimensional latent representation of the data can be used for visualization and
clustering.

TotalVI takes as input a scRNA-seq gene expression and protein expression matrix with cells and
genes.
We provide an extensive [user guide](https://docs.scvi-tools.org/en/1.2.0/user_guide/models/totalvi.html).

- See our original manuscript for further details of the model:
[TotalVI manuscript](https://www.nature.com/articles/s41592-020-01050-x).
- See our manuscript on [scvi-hub](https://www.biorxiv.org/content/10.1101/2024.03.01.582887v2)
how to leverage pre-trained models.

This model can be used for fine tuning on new data using our Arches framework:
[Arches tutorial](https://docs.scvi-tools.org/en/1.0.0/tutorials/notebooks/scarches_scvi_tools.html).


# Model Description

CITE-seq to measure RNA and surface proteins in thymocytes from wild-type and T cell lineage-restricted mice to generate a comprehensive timeline of cell state for each T cell lineage.

# Metrics

We provide here key performance metrics for the uploaded model, if provided by the data uploader.

<details>
<summary><strong>Coefficient of variation</strong></summary>

The cell-wise coefficient of variation summarizes how well variation between different cells is
preserved by the generated model expression. Below a squared Pearson correlation coefficient of 0.4
, we would recommend not to use generated data for downstream analysis, while the generated latent
space might still be useful for analysis.

**Cell-wise Coefficient of Variation**:

Modality: rna

| Metric                  | Training Value | Validation Value |
|-------------------------|----------------|------------------|
| Mean Absolute Error | 0.57  | 0.56           |
| Pearson Correlation | 0.76  | 0.76  |
| Spearman Correlation | 0.83 | 0.83  |
| R² (R-Squared) | -0.10  | -0.08      |

Modality: protein

| Metric                  | Training Value | Validation Value |
|-------------------------|----------------|------------------|
| Mean Absolute Error | 0.32  | 0.32           |
| Pearson Correlation | 0.53  | 0.53  |
| Spearman Correlation | 0.78 | 0.78  |
| R² (R-Squared) | -1.46  | -1.43      |



The gene-wise coefficient of variation summarizes how well variation between different genes is
preserved by the generated model expression. This value is usually quite high.

**Gene-wise Coefficient of Variation**:

Modality: rna

| Metric                  | Training Value |
|-------------------------|----------------|
| Mean Absolute Error | 26.96   |
| Pearson Correlation | 0.95  |
| Spearman Correlation | 0.99 |
| R² (R-Squared) | -0.25  |

Modality: protein

| Metric                  | Training Value |
|-------------------------|----------------|
| Mean Absolute Error | 4.30   |
| Pearson Correlation | 0.40  |
| Spearman Correlation | 0.73 |
| R² (R-Squared) | -6.19  |



</details>

<details>
<summary><strong>Differential expression metric</strong></summary>

The differential expression metric provides a summary of the differential expression analysis
between cell types or input clusters. We provide here the F1-score, Pearson Correlation
Coefficient of Log-Foldchanges, Spearman Correlation Coefficient, and Area Under the Precision
Recall Curve (AUPRC) for the differential expression analysis using Wilcoxon Rank Sum test for each
cell-type.

**Differential expression**:

Modality: rna

| Index | gene_f1 | lfc_mae | lfc_pearson | lfc_spearman | roc_auc | pr_auc | n_cells |
| --- | --- | --- | --- | --- | --- | --- | --- |
| CD14-positive monocyte | 0.95 | 2.11 | 0.59 | 0.91 | 0.09 | 0.02 | 120843.00 |
| CD16-positive, CD56-dim natural killer cell, human | 0.95 | 2.35 | 0.45 | 0.84 | 0.09 | 0.02 | 92848.00 |
| naive thymus-derived CD4-positive, alpha-beta T cell | 0.89 | 2.76 | 0.39 | 0.75 | 0.09 | 0.02 | 63096.00 |
| effector CD8-positive, alpha-beta T cell | 0.88 | 3.49 | 0.40 | 0.72 | 0.07 | 0.02 | 53534.00 |
| central memory CD4-positive, alpha-beta T cell | 0.93 | 2.59 | 0.34 | 0.74 | 0.06 | 0.02 | 49904.00 |
| naive B cell | 0.93 | 3.29 | 0.40 | 0.72 | 0.08 | 0.02 | 44136.00 |
| naive thymus-derived CD8-positive, alpha-beta T cell | 0.93 | 3.54 | 0.37 | 0.67 | 0.07 | 0.02 | 31175.00 |
| mature NK T cell | 0.91 | 3.62 | 0.44 | 0.63 | 0.04 | 0.01 | 21673.00 |
| effector memory CD8-positive, alpha-beta T cell | 0.82 | 4.47 | 0.37 | 0.56 | 0.07 | 0.02 | 18917.00 |
| T-helper 22 cell | 0.90 | 4.01 | 0.42 | 0.60 | 0.06 | 0.02 | 18379.00 |
| gamma-delta T cell | 0.88 | 4.52 | 0.39 | 0.50 | 0.05 | 0.01 | 15942.00 |
| platelet | 0.89 | 4.32 | 0.54 | 0.67 | 0.06 | 0.02 | 15847.00 |
| T follicular helper cell | 0.93 | 4.43 | 0.41 | 0.55 | 0.06 | 0.02 | 13608.00 |
| mucosal invariant T cell | 0.86 | 4.85 | 0.42 | 0.48 | 0.06 | 0.02 | 10992.00 |
| CD16-negative, CD56-bright natural killer cell, human | 0.85 | 5.29 | 0.38 | 0.44 | 0.05 | 0.02 | 10442.00 |
| class switched memory B cell | 0.89 | 5.17 | 0.45 | 0.49 | 0.08 | 0.02 | 7244.00 |
| immature B cell | 0.89 | 5.66 | 0.45 | 0.45 | 0.10 | 0.02 | 5238.00 |
| natural killer cell | 0.88 | 5.20 | 0.46 | 0.45 | 0.09 | 0.02 | 4963.00 |
| plasmacytoid dendritic cell | 0.90 | 5.18 | 0.46 | 0.46 | 0.05 | 0.02 | 4612.00 |
| CD14-low, CD16-positive monocyte | 0.91 | 4.38 | 0.58 | 0.59 | 0.10 | 0.02 | 4140.00 |
| plasmablast | 0.80 | 4.96 | 0.52 | 0.55 | 0.10 | 0.02 | 4121.00 |
| IgG plasma cell | 0.70 | 5.01 | 0.51 | 0.52 | 0.12 | 0.01 | 3527.00 |
| dendritic cell, human | 0.83 | 5.35 | 0.41 | 0.40 | 0.62 | 0.20 | 3357.00 |
| unswitched memory B cell | 0.92 | 5.08 | 0.49 | 0.47 | 0.17 | 0.02 | 3285.00 |
| myeloid dendritic cell | 0.82 | 5.44 | 0.47 | 0.45 | 0.13 | 0.02 | 3243.00 |
| B cell | 0.86 | 5.03 | 0.51 | 0.48 | 0.14 | 0.02 | 3024.00 |
| IgA plasma cell | 0.70 | 5.22 | 0.50 | 0.49 | 0.13 | 0.02 | 2699.00 |
| effector memory CD4-positive, alpha-beta T cell | 0.88 | 5.17 | 0.48 | 0.41 | 0.14 | 0.02 | 2634.00 |
| malignant cell | 0.94 | 5.40 | 0.48 | 0.45 | 0.28 | 0.02 | 2291.00 |
| CD34-positive, CD38-negative hematopoietic stem cell | 0.78 | 5.68 | 0.46 | 0.47 | 0.12 | 0.02 | 2238.00 |
| erythrocyte | 0.79 | 5.58 | 0.41 | 0.26 | 0.63 | 0.33 | 2232.00 |
| CD8-positive, alpha-beta T cell | 0.84 | 5.28 | 0.48 | 0.39 | 0.35 | 0.03 | 1355.00 |
| IgM plasma cell | 0.78 | 4.70 | 0.53 | 0.50 | 0.30 | 0.02 | 1163.00 |
| ILC1, human | 0.79 | 4.52 | 0.52 | 0.47 | 0.36 | 0.03 | 776.00 |
| erythroid progenitor cell, mammalian | 0.73 | 5.52 | 0.51 | 0.46 | 0.32 | 0.02 | 773.00 |
| monocyte | 0.89 | 4.21 | 0.57 | 0.51 | 0.38 | 0.02 | 649.00 |
| CD4-positive, alpha-beta T cell | 0.81 | 4.64 | 0.54 | 0.48 | 0.33 | 0.02 | 624.00 |
| dendritic cell | 0.75 | 5.15 | 0.50 | 0.41 | 0.47 | 0.03 | 585.00 |
| T-helper 1 cell | 0.86 | 4.12 | 0.55 | 0.51 | 0.35 | 0.02 | 481.00 |
| regulatory T cell | 0.77 | 3.97 | 0.56 | 0.54 | 0.36 | 0.02 | 329.00 |
| hematopoietic precursor cell | 0.68 | 4.34 | 0.60 | 0.57 | 0.30 | 0.02 | 180.00 |
| group 2 innate lymphoid cell, human | 0.66 | 3.08 | 0.58 | 0.63 | 0.26 | 0.02 | 93.00 |
| T-helper 2 cell | 0.65 | 2.71 | 0.62 | 0.67 | 0.18 | 0.02 | 55.00 |
| myeloid lineage restricted progenitor cell | 0.57 | 3.85 | 0.54 | 0.57 | 0.33 | 0.02 | 53.00 |
| megakaryocyte | 0.64 | 3.73 | 0.61 | 0.59 | 0.30 | 0.02 | 53.00 |
| T-helper 17 cell | 0.53 | 2.45 | 0.55 | 0.70 | 0.20 | 0.02 | 13.00 |

Modality: protein

| Index | gene_f1 | lfc_mae | lfc_pearson | lfc_spearman | roc_auc | pr_auc | n_cells |
| --- | --- | --- | --- | --- | --- | --- | --- |
| CD14-positive monocyte | 0.95 | 0.07 | 1.00 | 0.99 | 0.26 | 0.12 | 120843.00 |
| CD16-positive, CD56-dim natural killer cell, human | 0.95 | 0.06 | 0.99 | 0.98 | 0.22 | 0.12 | 92848.00 |
| naive thymus-derived CD4-positive, alpha-beta T cell | 0.84 | 0.06 | 0.99 | 0.98 | 0.37 | 0.13 | 63096.00 |
| effector CD8-positive, alpha-beta T cell | 0.95 | 0.05 | 0.99 | 0.97 | 0.17 | 0.09 | 53534.00 |
| central memory CD4-positive, alpha-beta T cell | 1.00 | 0.06 | 1.00 | 0.99 | 0.33 | 0.12 | 49904.00 |
| naive B cell | 1.00 | 0.08 | 1.00 | 0.96 | 0.20 | 0.12 | 44136.00 |
| naive thymus-derived CD8-positive, alpha-beta T cell | 0.95 | 0.07 | 0.99 | 0.95 | 0.13 | 0.08 | 31175.00 |
| mature NK T cell | 0.95 | 0.06 | 0.99 | 0.97 | 0.21 | 0.10 | 21673.00 |
| effector memory CD8-positive, alpha-beta T cell | 0.84 | 0.05 | 0.99 | 0.98 | 0.06 | 0.11 | 18917.00 |
| T-helper 22 cell | 0.95 | 0.06 | 0.99 | 0.98 | 0.11 | 0.08 | 18379.00 |
| gamma-delta T cell | 0.89 | 0.07 | 0.97 | 0.93 | 0.26 | 0.18 | 15942.00 |
| platelet | 0.79 | 0.10 | 0.97 | 0.95 | 0.21 | 0.11 | 15847.00 |
| T follicular helper cell | 1.00 | 0.07 | 1.00 | 0.99 | 0.24 | 0.12 | 13608.00 |
| mucosal invariant T cell | 0.89 | 0.08 | 0.97 | 0.94 | 0.15 | 0.09 | 10992.00 |
| CD16-negative, CD56-bright natural killer cell, human | 0.95 | 0.08 | 0.98 | 0.93 | 0.44 | 0.47 | 10442.00 |
| class switched memory B cell | 0.89 | 0.09 | 0.99 | 0.96 | 0.11 | 0.12 | 7244.00 |
| immature B cell | 0.89 | 0.13 | 0.98 | 0.93 | 0.26 | 0.14 | 5238.00 |
| natural killer cell | 0.89 | 0.06 | 0.98 | 0.98 | 0.68 | 0.70 | 4963.00 |
| plasmacytoid dendritic cell | 0.84 | 0.09 | 0.98 | 0.97 | 0.54 | 0.56 | 4612.00 |
| CD14-low, CD16-positive monocyte | 0.89 | 0.08 | 0.99 | 0.98 | 0.58 | 0.23 | 4140.00 |
| plasmablast | 0.79 | 0.08 | 0.99 | 0.97 | 0.47 | 0.49 | 4121.00 |
| IgG plasma cell | 0.89 | 0.08 | 0.99 | 0.95 | 0.47 | 0.51 | 3527.00 |
| dendritic cell, human | 0.79 | 0.10 | 0.97 | 0.90 | 0.94 | 0.90 | 3357.00 |
| unswitched memory B cell | 0.89 | 0.10 | 0.99 | 0.96 | 0.63 | 0.61 | 3285.00 |
| myeloid dendritic cell | 0.89 | 0.11 | 0.97 | 0.95 | 0.74 | 0.74 | 3243.00 |
| B cell | 0.89 | 0.10 | 0.98 | 0.92 | 0.58 | 0.59 | 3024.00 |
| IgA plasma cell | 0.89 | 0.10 | 0.97 | 0.91 | 0.47 | 0.48 | 2699.00 |
| effector memory CD4-positive, alpha-beta T cell | 0.95 | 0.08 | 0.99 | 0.95 | 0.79 | 0.80 | 2634.00 |
| malignant cell | 0.89 | 0.09 | 0.99 | 0.99 | 0.17 | 0.08 | 2291.00 |
| CD34-positive, CD38-negative hematopoietic stem cell | 0.84 | 0.09 | 0.97 | 0.95 | 0.37 | 0.35 | 2238.00 |
| erythrocyte | 0.89 | 0.07 | 0.99 | 0.98 | 0.21 | 0.25 | 2232.00 |
| CD8-positive, alpha-beta T cell | 0.63 | 0.09 | 0.92 | 0.87 | 0.68 | 0.52 | 1355.00 |
| IgM plasma cell | 0.95 | 0.09 | 0.98 | 0.92 | 0.42 | 0.45 | 1163.00 |
| ILC1, human | 0.84 | 0.10 | 0.97 | 0.89 | 0.53 | 0.55 | 776.00 |
| erythroid progenitor cell, mammalian | 0.68 | 0.14 | 0.95 | 0.93 | 0.23 | 0.23 | 773.00 |
| monocyte | 0.84 | 0.09 | 0.97 | 0.96 | 0.68 | 0.70 | 649.00 |
| CD4-positive, alpha-beta T cell | 0.84 | 0.12 | 0.91 | 0.85 | 0.58 | 0.59 | 624.00 |
| dendritic cell | 0.74 | 0.19 | 0.85 | 0.81 | 0.53 | 0.44 | 585.00 |
| T-helper 1 cell | 0.95 | 0.08 | 0.99 | 0.96 | 0.73 | 0.71 | 481.00 |
| regulatory T cell | 0.84 | 0.14 | 0.97 | 0.94 | 0.89 | 0.84 | 329.00 |
| hematopoietic precursor cell | 0.74 | 0.15 | 0.95 | 0.91 | 0.36 | 0.28 | 180.00 |
| group 2 innate lymphoid cell, human | 0.63 | 0.36 | 0.14 | 0.60 | 0.29 | 0.32 | 93.00 |
| T-helper 2 cell | 0.79 | 0.40 | 0.15 | 0.69 | 0.74 | 0.65 | 55.00 |
| myeloid lineage restricted progenitor cell | 0.53 | 0.28 | 0.97 | 0.74 | 0.56 | 0.37 | 53.00 |
| megakaryocyte | 0.47 | 0.25 | 0.98 | 0.79 | 0.54 | 0.44 | 53.00 |
| T-helper 17 cell | 0.79 | 0.55 | 0.82 | 0.78 | 0.48 | 0.38 | 13.00 |



</details>

# Model Properties

We provide here key parameters used to setup and train the model.

<details>
<summary><strong>Model Parameters</strong></summary>

These provide the settings to setup the original model:
```json
{
    "n_latent": 20,
    "gene_dispersion": "gene",
    "protein_dispersion": "protein",
    "gene_likelihood": "nb",
    "latent_distribution": "normal",
    "empirical_protein_background_prior": null,
    "override_missing_proteins": false
}
```

</details>

<details>
<summary><strong>Setup Data Arguments</strong></summary>

Arguments passed to setup_anndata of the original model:
```json
{
    "rna_layer": "counts",
    "protein_layer": null,
    "batch_key": "donor_id",
    "size_factor_key": null,
    "categorical_covariate_keys": null,
    "continuous_covariate_keys": null,
    "modalities": {
        "rna_layer": "rna",
        "protein_layer": "protein",
        "batch_key": "rna"
    }
}
```

</details>

<details>
<summary><strong>Data Registry</strong></summary>

Registry elements for AnnData manager:
| Registry Key             | scvi-tools Location                  |
|--------------------------|--------------------------------------|
| X                         | adata.mod['rna'].layers['counts']    |
| batch                     | adata.mod['rna'].obs['_scvi_batch']  |
| labels                    | adata.obs['_scvi_labels']            |
| latent_qzm                | adata.obsm['totalvi_latent_qzm']     |
| latent_qzv                | adata.obsm['totalvi_latent_qzv']     |
| minify_type               | adata.uns['_scvi_adata_minify_type'] |
| observed_lib_size         | adata.obs['observed_lib_size']       |
| proteins                  | adata.mod['protein'].X               |

- **Data is Minified**: False

</details>

<details>
<summary><strong>Summary Statistics</strong></summary>

| Summary Stat Key          | Value |
|--------------------------|-------|
| n_batch                   | 120 |
| n_cells                   | 647366 |
| n_extra_categorical_covs  | 0 |
| n_extra_continuous_covs   | 0 |
| n_labels                  | 1 |
| n_latent_qzm              | 20 |
| n_latent_qzv              | 20 |
| n_proteins                | 192 |
| n_vars                    | 4000 |

</details>


<details>
<summary><strong>Training</strong></summary>

<!-- If your model is not uploaded with any data (e.g., minified data) on the Model Hub, then make
sure to provide this field if you want users to be able to access your training data. See the
scvi-tools documentation for details. -->
**Training data url**: Not provided by uploader

If provided by the original uploader, for those interested in understanding or replicating the
training process, the code is available at the link below.

**Training Code URL**: https://github.com/YosefLab/Thymus_CITE-seq/blob/main/totalVI_AllData/totalVI_thymus111.ipynb

</details>


# References

Steier, Z., Aylard, D.A., McIntyre, L.L. et al. Single-cell multiomic analysis of thymocyte development reveals drivers of CD4+ T cell and CD8+ T cell lineage commitment. Nat Immunol 24, 1579–1590 (2023). https://doi.org/10.1038/s41590-023-01584-0.