Related Experiment Video
Updated: Jun 6, 2025
10:12
Droplet Barcoding-Based Single Cell Transcriptomics of Adult Mammalian Tissues
Published on: January 10, 2019
18.5K
A deconvolution framework that uses single-cell sequencing plus a small benchmark data set for accurate analysis of
Shuai Guo1, Xiaoqian Liu1, Xuesen Cheng2
1Department of Bioinformatics and Computational Biology, The University of Texas MD Anderson Cancer Center, Houston, Texas 77030, USA.
Genome Research
|November 25, 2024
Summary
DeMixSC improves bulk deconvolution accuracy by addressing technological discrepancies in single-cell RNA sequencing data. This new framework enables precise analysis of complex biological samples and disease tissues.
Area of Science:
- Genomics
- Bioinformatics
- Computational Biology
Background:
- Bulk deconvolution using single-cell/nucleus RNA sequencing (sc/nRNA-seq) is vital for analyzing biological sample heterogeneity.
- Technological variations between sequencing platforms hinder deconvolution accuracy.
Purpose of the Study:
- To develop a deconvolution framework, DeMixSC, that overcomes technological discrepancies in sc/nRNA-seq data.
- To improve the accuracy of cell type proportion estimation in complex tissues.
Main Methods:
- Utilized an experimental design to match inter-platform biological signals and identify technological discrepancies.
- Developed DeMixSC, a weighted nonnegative least-squares framework, to adjust for gene-wise technological differences.
- Applied DeMixSC to benchmark datasets (retinas, ovarian cancer) and large patient cohorts (age-related macular degeneration, ovarian cancer).
Main Results:
- DeMixSC demonstrated significantly improved deconvolution accuracy on benchmark datasets.
- The framework successfully identified biologically meaningful differences in patient cohorts, including age-related macular degeneration and ovarian cancer.
- DeMixSC outperformed existing methods in large-scale deconvolution tasks.
Conclusions:
- Technological discrepancy significantly impacts deconvolution performance.
- A well-matched benchmark dataset is crucial for accurate deconvolution.
- DeMixSC offers a broadly applicable solution for accurate deconvolution of disease tissues in large patient cohorts.
