Does it support batch correction across different labs?

Yes, scvi-tools is specifically designed to handle technical variation and batch effects by registering study-specific covariates during model setup.

Can I use this for multi-modal data integration?

Yes, it supports multimodal integration including CITE-seq (totalVI), multiome (MultiVI), and spatial transcriptomics deconvolution (DestVI).

How does it interact with the Scanpy ecosystem?

It follows a unified API where results like latent representations and normalized values are stored directly back into AnnData objects for further downstream analysis in Scanpy.

Is GPU acceleration supported for large datasets?

Absolutely; the skill provides guidance on utilizing PyTorch-based GPU acceleration to significantly speed up model training on large-scale datasets.

What data format does this skill require?

The scvi-tools skill is designed to work with AnnData objects and requires raw, unnormalized count data for optimal probabilistic modeling.

SCVI Tools for Single-Cell Omics

Name: SCVI Tools for Single-Cell Omics
Author: Microck

byMicrock

•

Ciencia de Datos y ML

Analyzes single-cell omics data using deep generative models for batch correction, integration, and differential expression.

This skill empowers Claude to perform advanced single-cell genomics analysis using the scvi-tools framework, a Python library for probabilistic modeling. It facilitates the implementation of deep generative models across diverse modalities—including scRNA-seq, scATAC-seq, and spatial transcriptomics—enabling researchers to handle technical batch effects, integrate multimodal datasets, and perform statistically rigorous differential expression analysis within a unified PyTorch-based ecosystem. It is particularly useful for scientists and bioinformaticians looking to leverage variational inference for high-dimensional biological data.

Características Principales

01Seamless integration with the Scanpy and AnnData ecosystem

02Multi-modal support for CITE-seq, multiome, and spatial transcriptomics

03Deep generative modeling for single-cell data integration and batch correction

04Probabilistic differential expression and uncertainty quantification

05Semi-supervised cell type annotation and transfer learning

0681 GitHub stars

Casos de Uso

01Integrating multiple scRNA-seq datasets across different studies and donors

02Performing spatial deconvolution and mapping for spatial transcriptomics data

03Jointly analyzing protein and gene expression data from CITE-seq experiments

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add microck/ordinary-claude-skills scvi-tools

For use in Claude.ai and ChatGPT

Download Skill