About
This skill provides specialized support for scvi-tools, a Python framework for probabilistic modeling of single-cell genomics data. It assists users in implementing deep learning models like scVI and scANVI to handle complex tasks such as batch correction, multimodal integration, and spatial transcriptomics analysis. By leveraging the AnnData ecosystem and PyTorch-based inference, it guides developers and researchers through the entire workflow—from data registration and model training to downstream analysis and uncertainty quantification—ensuring best practices like using raw counts and proper covariate handling are followed.
