Related Experiment Video
Updated: Jun 11, 2025
08:51
Author Spotlight: Integrated Multi-Omics Analysis for Unveiling Multicellular Immune Signatures in Clinical Heart Attack Cohorts
Published on: September 20, 2024
1.2K
MultiSC: a deep learning pipeline for analyzing multiomics single-cell data
Xiang Lin1,2, Siqi Jiang2, Le Gao1,2
1Department of Quantitative Health Sciences, Mayo Clinic, 13400 E Shea Blvd, Scottsdale, AZ 85259, United States.
Briefings in Bioinformatics
|October 8, 2024
Summary
Researchers developed MultiSC, a novel pipeline for analyzing multi-omics single-cell data. This tool integrates gene expression, chromatin accessibility, and transcription factor protein data to reveal cell activities and gene regulatory networks.
Area of Science:
- Single-cell biology
- Genomics
- Proteomics
- Systems biology
Background:
- Single-cell technologies offer high-resolution insights into cellular functions and processes.
- Multi-omics single-cell sequencing techniques capture diverse cellular behaviors.
- NEAT-seq simultaneously generates gene expression, chromatin accessibility, and transcription factor protein data, but integration tools are lacking.
Purpose of the Study:
- To develop a novel computational pipeline, MultiSC, for the integrated analysis of multi-omics single-cell data.
- To address the gap in tools for effectively integrating gene expression, chromatin accessibility, and transcription factor protein data.
- To provide a comprehensive framework for understanding cellular activities and gene regulation.
Main Methods:
- MultiSC pipeline utilizes a multimodal constraint autoencoder for data integration during clustering.
- Employs a matrix factorization-based model (scMF) for predicting transcription factor target genes.
- Incorporates multivariate linear regression for gene regulatory network prediction.
- Includes functionalities for differential expression, mediation analysis, and causal inference.
Main Results:
- The MultiSC pipeline effectively integrates multi-omics single-cell data, including gene expression, chromatin accessibility, and TF protein levels.
- Demonstrated ability to provide a comprehensive view of cellular activities and gene regulatory networks.
- Experimental evaluations confirmed the pipeline's performance in leveraging multi-omics data.
Conclusions:
- MultiSC enhances the understanding of cellular processes by enabling effective integration and analysis of diverse omics data types.
- The pipeline empowers researchers to fully leverage the potential of multi-omics single-cell data for biological discovery.
- Facilitates deeper insights into cell behavior and regulatory mechanisms through integrated multi-modal analysis.
