---
library_name: scvi-tools
license: cc-by-4.0
tags:
- biology
- genomics
- single-cell
- model_cls_name:SCVI
- scvi_version:1.4.2
- anndata_version:0.12.7
- modality:rna
- tissue:various
- annotated:True
---


ScVI is a variational inference model for single-cell RNA-seq data that can learn an underlying
latent space, integrate technical batches and impute dropouts.
The learned low-dimensional latent representation of the data can be used for visualization and
clustering.

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

- See our original manuscript for further details of the model:
[scVI manuscript](https://www.nature.com/articles/s41592-018-0229-2).
- 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/stable/tutorials/notebooks/scrna/scarches_scvi_tools.html).


# Model Description

Tabula Sapiens is a benchmark, first-draft human cell atlas of nearly 500,000 cells from 24 organs of 15 normal human subjects.

# 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**:

| Metric                  | Training Value | Validation Value |
|-------------------------|----------------|------------------|
| Mean Absolute Error | 1.68  | 1.79           |
| Pearson Correlation | 0.75  | 0.72  |
| Spearman Correlation | 0.74 | 0.72  |
| R² (R-Squared) | 0.30  | 0.19      |

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**:

| Metric                  | Training Value |
|-------------------------|----------------|
| Mean Absolute Error | 19.82   |
| Pearson Correlation | 0.69  |
| Spearman Correlation | 0.74 |
| R² (R-Squared) | -0.52  |

</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**:

| Index | gene_f1 | lfc_mae | lfc_pearson | lfc_spearman | roc_auc | pr_auc | n_cells |
| --- | --- | --- | --- | --- | --- | --- | --- |
| luminal epithelial cell of mammary gland | 0.93 | 0.22 | 0.94 | 0.99 | 0.29 | 0.10 | 15432.00 |
| fibroblast of breast | 0.95 | 0.35 | 0.85 | 0.97 | 0.14 | 0.03 | 5270.00 |
| CD8-positive, alpha-beta T cell | 0.91 | 1.23 | 0.59 | 0.91 | 0.03 | 0.02 | 2653.00 |
| basal cell | 0.90 | 0.68 | 0.68 | 0.93 | 0.38 | 0.38 | 1819.00 |
| progenitor cell of mammary luminal epithelium | 0.88 | 0.68 | 0.61 | 0.91 | 0.90 | 0.85 | 1694.00 |
| endothelial cell | 0.90 | 1.03 | 0.58 | 0.91 | 0.47 | 0.47 | 1534.00 |
| macrophage | 0.87 | 1.33 | 0.61 | 0.89 | 0.44 | 0.40 | 685.00 |
| CD4-positive, alpha-beta T cell | 0.91 | 2.49 | 0.56 | 0.80 | 0.15 | 0.12 | 681.00 |
| vascular associated smooth muscle cell | 0.88 | 1.81 | 0.59 | 0.85 | 0.37 | 0.34 | 566.00 |
| plasma cell | 0.80 | 2.06 | 0.64 | 0.80 | 0.09 | 0.02 | 293.00 |
| T cell | 0.84 | 4.03 | 0.47 | 0.59 | 0.40 | 0.05 | 223.00 |
| mast cell | 0.70 | 6.44 | 0.49 | 0.45 | 0.35 | 0.02 | 37.00 |
| B cell | 0.49 | 7.59 | 0.44 | 0.41 | 0.30 | 0.02 | 15.00 |
| mature NK T cell | 0.48 | 6.97 | 0.45 | 0.42 | 0.35 | 0.03 | 14.00 |
| basophil | 0.51 | 6.45 | 0.43 | 0.37 | 0.34 | 0.03 | 12.00 |
| natural killer cell | 0.38 | 7.13 | 0.37 | 0.37 | 0.40 | 0.03 | 6.00 |
| regulatory T cell | 0.27 | 4.74 | 0.32 | 0.51 | 0.35 | 0.03 | 2.00 |

</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_hidden": 128,
    "n_latent": 20,
    "n_layers": 3,
    "dropout_rate": 0.05,
    "dispersion": "gene",
    "gene_likelihood": "nb",
    "use_observed_lib_size": true,
    "latent_distribution": "normal",
    "use_batch_norm": "none",
    "use_layer_norm": "both",
    "encode_covariates": true
}
```

</details>

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

Arguments passed to setup_anndata of the original model:
```json
{
    "layer": "counts",
    "batch_key": "donor_assay",
    "labels_key": "cell_type",
    "size_factor_key": null,
    "categorical_covariate_keys": null,
    "continuous_covariate_keys": null
}
```

</details>

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

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

- **Data is Minified**: False

</details>

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

| Summary Stat Key          | Value |
|--------------------------|-------|
| n_batch                   | 3 |
| n_cells                   | 30936 |
| n_extra_categorical_covs  | 0 |
| n_extra_continuous_covs   | 0 |
| n_labels                  | 17 |
| n_latent_qzm              | 20 |
| n_latent_qzv              | 20 |
| n_vars                    | 3000 |

</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/scvi-hub-models/blob/main/src/scvi_hub_models/TS_train_all_tissues.ipynb

</details>


# References

The Tabula Sapiens Consortium. The Tabula Sapiens: A multiple-organ, single-cell transcriptomic atlas of humans. Science, May 2022. doi:10.1126/science.abl4896