Related Experiment Video
Updated: Jul 23, 2026
10:12
Droplet Barcoding-Based Single Cell Transcriptomics of Adult Mammalian Tissues
Published on: January 10, 2019
18.9K
GMM-Demux: sample demultiplexing, multiplet detection, experiment planning, and novel cell-type verification in
Hongyi Xin1,2, Qiuyu Lian2,3, Yale Jiang2,4
1University of Michigan-Shanghai Jiao Tong University Joint Institute, Shanghai Jiao Tong University, Shanghai, 200240, China.
Genome Biology
|August 1, 2020
Summary
Multiplets in single-cell RNA sequencing (scRNA-seq) create fake cell types. Our new method, GMM-Demux, uses Gaussian mixture models and sample barcoding to accurately identify and remove these multiplets, improving data reliability.
Area of Science:
- Genomics and Bioinformatics
- Computational Biology
- Molecular Biology
Background:
- Single-cell RNA sequencing (scRNA-seq) is crucial for understanding cellular heterogeneity.
- Multiplets, where multiple cells are captured in a single droplet, are a significant challenge in scRNA-seq, leading to artificial cell types and reduced data accuracy.
- Accurate identification and removal of multiplets are essential for reliable scRNA-seq data analysis and interpretation.
Purpose of the Study:
- To develop and validate a novel computational method for identifying and removing multiplets in scRNA-seq data.
- To improve the scalability and reliability of scRNA-seq analyses by addressing the issue of multiplet contamination.
- To authenticate putative cell types identified in scRNA-seq datasets by accounting for potential multiplet-induced artifacts.
Main Methods:
- Proposed GMM-Demux, a Gaussian mixture model-based method for multiplet identification.
- Utilized sample barcoding techniques, including cell hashing and MULTI-seq, for multiplet detection.
- Incorporated a droplet formation model within GMM-Demux to authenticate cell types derived from scRNA-seq data.
- Generated two in-house cell-hashing datasets for method evaluation.
- Compared GMM-Demux performance against three existing state-of-the-art sample barcoding classifiers.
Main Results:
- GMM-Demux demonstrated high accuracy and stability in identifying and removing multiplets.
- The method successfully recognized 9 multiplet-induced artificial cell types within a peripheral blood mononuclear cell (PBMC) dataset.
- Comparative analysis showed GMM-Demux's superior performance compared to other leading classifiers.
Conclusions:
- GMM-Demux provides a robust and accurate solution for multiplet identification and removal in scRNA-seq data.
- The proposed method enhances the reliability and scalability of scRNA-seq analyses.
- GMM-Demux effectively mitigates the impact of multiplets, leading to more trustworthy biological insights from scRNA-seq experiments.
