Model single-cell omics data with probabilistic, batch-corrected latent representations.
npx skills add https://github.com/swaruplab/operon --skill scvi-tools-swaruplab
Please help me install this Agent Skill. Skill: scvi-tools Source: https://github.com/swaruplab/operon/tree/main/src-tauri/protocols/scvi-tools Command: npx skills add https://github.com/swaruplab/operon --skill scvi-tools-swaruplab
scvi-tools provides deep generative models for single-cell omics, enabling probabilistic batch correction, transfer learning, and multi-modal integration to extract robust biological signals from noisy data. This suite supports advanced analysis across RNA, ATAC, and protein modalities, delivering batch-invariant latent representations and principled differential expression with uncertainty.
Train a scvi-tools model on your single-cell dataset to obtain a latent, batch-corrected representation and perform differential expression.
High-intent search queries and answers about installing and using this skill.
Probabilistic batch correction for single-cell RNA data uses deep generative models to produce batch-invariant latent representations. This requires an AnnData object containing raw counts and batch metadata to train the model and extract corrected embeddings.
Differential expression with uncertainty is supported directly on AnnData objects by leveraging Bayesian variational inference. This approach provides principled statistical bounds on gene expression differences rather than simple point estimates.
Cross-modality integration for RNA and ATAC data is supported using the MultiVI model. It enables scalable inference on AnnData objects to extract robust biological signals and integrate diverse omics modalities.
Integrating multimodal single-cell omics datasets is best achieved using deep generative models like TOTALVI. These models handle RNA, ATAC, and protein data to deliver batch-invariant latent representations and transfer learning across modalities.
Raw counts are strictly required in your AnnData object to train single-cell probabilistic models. Normalized data is not suitable for the Bayesian variational inference process that underlies batch correction and latent representation generation.
Probabilistic models provide principled uncertainty estimates for differential expression and robust batch-invariant representations. Standard methods often lack statistical bounds and struggle with noisy single-cell data and complex batch effects.
dralkhAnalyze single-cell omics data with probabilistic deep generative models.
HolobiomicsLabTrain scvi-tools models on AnnData for batch-corrected latent spaces and differential expression.
jackspaceModel single-cell omics data with scvi-tools for batch-corrected analysis.
LogauaEngstromAnalyze single-cell omics data with batch correction and cross-modality integration.
