bio-single-cell-multimodal-integration

Integrate multi-modal single-cell RNA, protein, and ATAC data into unified representations.

Updated Mar 30, 2026

Search Tags:#integration #multimodal #single-cell #scanpy #seurat #muon #wnn

Ready to run

npx skills add https://github.com/ya-way/cytoclaw-skills --skill bio-single-cell-multimodal-integration

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: bio-single-cell-multimodal-integration
Source: https://github.com/ya-way/cytoclaw-skills/tree/main/workspace/skills/bio-single-cell-multimodal-integration
Command: npx skills add https://github.com/ya-way/cytoclaw-skills --skill bio-single-cell-multimodal-integration

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Multimodal single-cell experiments measure several cellular layers (RNA, protein, chromatin), and integrating these modalities into a cohesive analysis is error-prone and time-consuming. This skill provides a unified framework to load, normalize, and jointly analyze multi-omics data, enabling consistent cell-type discovery and cross-modality interpretation.

Core Features & Use Cases

Multimodal integration: combine RNA with protein (ADT) or RNA with ATAC to build a joint representation.
Cross-platform workflows: supports Seurat-based (R) and MuData/Scanpy-based (Python) pipelines for end-to-end analysis.
Visualization & interpretation: generate joint UMAPs, modality weights, and cross-modality marker discovery for robust cell-type annotation.
Use Case: analyze CITE-seq or 10X Multiome datasets to identify common cell states across modalities and compare modality contributions.

Quick Start

Load your multimodal dataset (RNA + ADT or RNA + ATAC), run a WNN-based integration, and generate a joint clustering and visualization to start analyzing cell states.

Frequently Asked Questions about bio-single-cell-multimodal-integration

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I integrate multi-modal single-cell data for joint clustering and visualization?▼

To integrate multimodal single-cell data, apply modality-specific normalization (RNA log, ADT CLR, ATAC TF-IDF), use weighted nearest neighbors (WNN) for integration, and generate a joint UMAP for unified clustering and visualization.

What is the best way to analyze CITE-seq and 10X Multiome datasets across both R and Python?▼

Analyze CITE-seq and 10X Multiome datasets using cross-platform workflows that support Seurat in R and MuData/Scanpy in Python, enabling end-to-end loading, normalization, dimension reduction, and joint cell-state analysis.

Does weighted nearest neighbors work with both RNA and ATAC modalities for cell-type discovery?▼

Yes, weighted nearest neighbors (WNN) works with RNA and ATAC modalities by calculating modality weights, combining them into a unified cell representation, and enabling robust cross-modality marker discovery for cell-type annotation.

How do I normalize different modalities when integrating RNA and protein single-cell data?▼

Normalize different modalities by applying log normalization for RNA and centered log ratio (CLR) for protein (ADT) data before constructing a joint representation using weighted nearest neighbors.

Can I use Scanpy and MuData for single-cell multimodal integration instead of Seurat?▼

Yes, you can use Scanpy and MuData for single-cell multimodal integration as a Python alternative to Seurat, providing end-to-end workflows for loading, modality-specific normalization, dimension reduction, and joint UMAP visualization.

When should I adjust modality weights in a multi-omics single-cell integration workflow?▼

Adjust modality weights in a multi-omics integration workflow when comparing modality contributions to ensure consistent cell-type discovery and accurate cross-modality interpretation across RNA, protein, or chromatin data.