scrnaseq-deep-learning

What problem does it solve?

This Skill provides a deep learning framework for analyzing single-cell RNA-seq data, addressing the challenges of batch correction, label transfer, perturbation prediction, and foundation model fine-tuning.

Core Features & Use Cases

• Batch Correction & Integration: Utilize scVI and scANVI for batch effect removal and integration of scRNA-seq data.
• Label Transfer: Apply scANVI to predict cell types on new, unlabeled datasets based on a labeled reference dataset.
• Perturbation Prediction: Employ scVI and scGPT to predict perturbation effects from baseline data.
• Foundation Model Fine-tuning: Fine-tune scGPT and Geneformer models on user-specific datasets for downstream analysis.
• Use Case: Suppose you have scRNA-seq data from a perturbation experiment. Use this Skill to correct batch effects, predict perturbation effects, and annotate cell types.

Quick Start

Run the 'analyze-scrnaseq' script to correct batch effects and perform label transfer on your single-cell RNA-seq data.