scvi-tools Single-Cell Analysis

Acerca de

Provides a comprehensive framework for applying deep generative models and variational inference to single-cell omics data. This skill enables advanced workflows for scRNA-seq, scATAC-seq, and spatial transcriptomics by leveraging PyTorch-based models like scVI and totalVI. It streamlines complex tasks such as dimensionality reduction, batch correction across multiple studies, and multimodal integration, allowing researchers to extract meaningful biological insights from large-scale, high-dimensional datasets while maintaining statistical rigor through probabilistic modeling.

Características Principales

• Covariate registration for automated batch correction and technical noise handling
• Joint modeling of multimodal data types including RNA, protein, and chromatin
• Probabilistic differential expression analysis with uncertainty quantification
• Unified API for training deep generative models on single-cell count data
• 2,066 GitHub stars
• Spatially-resolved transcriptomics deconvolution and cell-type mapping

Casos de Uso

• Analyzing paired or unpaired multi-omic data such as CITE-seq and Multiome
• Mapping cell-type distributions in spatial transcriptomics using reference-based deconvolution
• Integrating disparate single-cell RNA-seq datasets while correcting for batch effects