--- library_name: scvi-tools license: cc-by-4.0 tags: - biology - genomics - single-cell - model_cls_name:CondSCVI - scvi_version:1.2.0 - anndata_version:0.11.1 - modality:rna - tissue:various - annotated:True --- CondSCVI is a variational inference model for single-cell RNA-seq data that can learn an underlying latent space. The predictions of the model are meant to be afterward used for deconvolution of a second spatial transcriptomics dataset in DestVI. DestVI predicts the cell-type proportions as well as cell type-specific activation state in the spatial data. CondSCVI takes as input a scRNA-seq gene expression matrix with cells and genes as well as a cell-type annotation for all cells. We provide an extensive [user guide](https://docs.scvi-tools.org/en/1.2.0/user_guide/models/destvi.html) for DestVI including a description of CondSCVI. - See our original manuscript for further details of the model: [DestVI manuscript](https://www.nature.com/articles/s41587-022-01272-8). - See our manuscript on [scvi-hub](https://www.biorxiv.org/content/10.1101/2024.03.01.582887v2) how to leverage pre-trained models. # 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.
Coefficient of variation 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 | 3.27 | 3.24 | | Pearson Correlation | 0.49 | 0.49 | | Spearman Correlation | 0.39 | 0.38 | | R² (R-Squared) | -1.58 | -1.55 | 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 | 24.57 | | Pearson Correlation | 0.73 | | Spearman Correlation | 0.84 | | R² (R-Squared) | -0.17 |
Differential expression metric 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 | | --- | --- | --- | --- | --- | --- | --- | --- | | macrophage | 0.91 | 2.03 | 0.61 | 0.85 | 0.29 | 0.93 | 12160.00 | | type II pneumocyte | 0.94 | 1.49 | 0.56 | 0.86 | 0.36 | 0.88 | 9146.00 | | basal cell | 0.76 | 2.04 | 0.54 | 0.70 | 0.44 | 0.81 | 2188.00 | | capillary endothelial cell | 0.88 | 3.46 | 0.50 | 0.65 | 0.30 | 0.80 | 1534.00 | | classical monocyte | 0.86 | 3.61 | 0.58 | 0.65 | 0.28 | 0.83 | 1487.00 | | club cell | 0.85 | 3.17 | 0.58 | 0.62 | 0.33 | 0.79 | 1038.00 | | non-classical monocyte | 0.78 | 3.58 | 0.60 | 0.65 | 0.33 | 0.84 | 1005.00 | | respiratory goblet cell | 0.91 | 3.00 | 0.60 | 0.67 | 0.39 | 0.82 | 762.00 | | basophil | 0.76 | 4.25 | 0.56 | 0.57 | 0.31 | 0.86 | 686.00 | | lung ciliated cell | 0.75 | 3.53 | 0.57 | 0.65 | 0.47 | 0.86 | 602.00 | | CD8-positive, alpha-beta T cell | 0.78 | 4.49 | 0.55 | 0.59 | 0.34 | 0.84 | 552.00 | | CD4-positive, alpha-beta T cell | 0.84 | 4.61 | 0.54 | 0.54 | 0.31 | 0.83 | 543.00 | | vein endothelial cell | 0.81 | 4.46 | 0.51 | 0.57 | 0.41 | 0.80 | 508.00 | | lung microvascular endothelial cell | 0.75 | 3.76 | 0.65 | 0.67 | 0.38 | 0.86 | 485.00 | | adventitial cell | 0.85 | 4.81 | 0.56 | 0.70 | 0.41 | 0.82 | 373.00 | | fibroblast | 0.89 | 4.75 | 0.50 | 0.57 | 0.38 | 0.78 | 336.00 | | dendritic cell | 0.86 | 4.87 | 0.54 | 0.60 | 0.38 | 0.83 | 316.00 | | intermediate monocyte | 0.74 | 4.25 | 0.62 | 0.62 | 0.38 | 0.85 | 252.00 | | pericyte | 0.86 | 4.97 | 0.55 | 0.53 | 0.39 | 0.79 | 213.00 | | type I pneumocyte | 0.81 | 4.02 | 0.63 | 0.64 | 0.38 | 0.79 | 211.00 | | endothelial cell of artery | 0.79 | 5.17 | 0.55 | 0.58 | 0.39 | 0.79 | 187.00 | | neutrophil | 0.83 | 5.12 | 0.55 | 0.50 | 0.37 | 0.90 | 174.00 | | plasma cell | 0.70 | 5.80 | 0.46 | 0.39 | 0.44 | 0.89 | 134.00 | | effector CD4-positive, alpha-beta T cell | 0.73 | 5.07 | 0.56 | 0.52 | 0.36 | 0.81 | 132.00 | | mature NK T cell | 0.66 | 5.31 | 0.53 | 0.51 | 0.40 | 0.80 | 132.00 | | B cell | 0.59 | 5.61 | 0.51 | 0.45 | 0.37 | 0.77 | 87.00 | | bronchial smooth muscle cell | 0.83 | 5.58 | 0.53 | 0.50 | 0.42 | 0.81 | 81.00 | | effector CD8-positive, alpha-beta T cell | 0.72 | 4.94 | 0.57 | 0.54 | 0.37 | 0.83 | 80.00 | | vascular associated smooth muscle cell | 0.82 | 4.97 | 0.58 | 0.53 | 0.45 | 0.81 | 80.00 | | endothelial cell of lymphatic vessel | 0.49 | 5.86 | 0.50 | 0.46 | 0.39 | 0.76 | 47.00 | | blood vessel endothelial cell | 0.67 | 5.04 | 0.59 | 0.56 | 0.41 | 0.82 | 47.00 | | smooth muscle cell | 0.63 | 4.92 | 0.58 | 0.53 | 0.38 | 0.81 | 25.00 | | pulmonary ionocyte | 0.18 | 6.58 | 0.42 | 0.31 | 0.38 | 0.72 | 19.00 | | plasmacytoid dendritic cell | 0.38 | 5.92 | 0.41 | 0.40 | 0.38 | 0.84 | 18.00 | | mesothelial cell | 0.38 | 5.69 | 0.54 | 0.45 | 0.38 | 0.72 | 17.00 | | serous cell of epithelium of bronchus | 0.40 | 5.99 | 0.47 | 0.38 | 0.41 | 0.80 | 15.00 |
# Model Properties We provide here key parameters used to setup and train the model.
Model Parameters These provide the settings to setup the original model: ```json { "n_hidden": 128, "n_latent": 5, "n_layers": 2, "weight_obs": false, "dropout_rate": 0.05 } ```
Setup Data Arguments Arguments passed to setup_anndata of the original model: ```json { "labels_key": "cell_ontology_class", "layer": null, "batch_key": null } ```
Data Registry Registry elements for AnnData manager: | Registry Key | scvi-tools Location | |--------------|---------------------------| | X | adata.X | | labels | adata.obs['_scvi_labels'] | - **Data is Minified**: False
Summary Statistics | Summary Stat Key | Value | |------------------|-------| | n_cells | 35672 | | n_labels | 36 | | n_vars | 3000 |
Training **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
# 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