Related Experiment Video
Updated: Jun 19, 2025
06:24
Multiplexed Analysis of Retinal Gene Expression and Chromatin Accessibility Using scRNA-Seq and scATAC-Seq
Published on: March 12, 2021
3.6K
CrossMP: Enabling Cross-Modality Translation between Single-Cell RNA-Seq and Single-Cell ATAC-Seq through Web-Based
Zhen Lyu1, Sabin Dahal1, Shuai Zeng1,2
1Department of Electrical Engineering and Computer Science, University of Missouri, Columbia, MO 65211, USA.
Genes
|July 27, 2024
Summary
Researchers developed a deep learning model to predict single-cell chromatin accessibility from gene expression data, and vice versa. This advances multiomic single-cell analysis, overcoming data generation challenges.
Area of Science:
- Computational Biology
- Genomics
- Molecular Biology
Background:
- Simultaneous profiling of multiple molecular data types (multiomic modalities) within single cells is crucial for a comprehensive understanding of cellular states.
- Integrating single-cell RNA sequencing (scRNA-seq) and single-cell transposase-accessible chromatin sequencing (scATAC-seq) offers deeper biological insights than single-modality analysis.
- Generating paired multiomic single-cell data is technically challenging and costly, limiting its widespread availability.
Purpose of the Study:
- To develop a computational model for cross-modal prediction between single-cell transcriptomic and chromatin accessibility profiles.
- To enable accurate prediction of one modality from another, mitigating the need for expensive paired data generation.
- To provide researchers with an accessible tool for multiomic data analysis.
Main Methods:
- A deep neural network architecture was employed to learn latent representations from source single-cell data modalities.
- The model translates between scRNA-seq and scATAC-seq data, predicting one modality from the other.
- The model's performance was validated across multiple paired human datasets.
Main Results:
- The proposed deep learning model demonstrated reliable performance in accurately translating between transcriptomic and chromatin accessibility profiles.
- The cross-modal prediction achieved high accuracy across diverse human single-cell datasets.
- A web-based portal, CrossMP, was developed for user-friendly access to the prediction model.
Conclusions:
- The developed model effectively bridges the gap between different single-cell omics data types, facilitating multiomic analysis.
- CrossMP provides a valuable resource for researchers to predict missing modalities, enhancing the utility of existing single-cell datasets.
- This approach democratizes multiomic data analysis, enabling broader scientific exploration without requiring paired data generation.
