Self-supervised contrastive learning for integrative single cell RNA-seq data analysis

Wenkai Han; Yuqi Cheng; Jiayang Chen; Huawen Zhong; Zhihang Hu; Siyuan Chen; Licheng Zong; Liang Hong; Ting Fung Chan; Irwin King; Xin Gao; Yu Li

doi:10.1093/bib/bbac377

Self-supervised contrastive learning for integrative single cell RNA-seq data analysis

Wenkai Han, Yuqi Cheng, Jiayang Chen, Huawen Zhong, Zhihang Hu, Siyuan Chen, Licheng Zong, Liang Hong, Ting Fung Chan, Irwin King, Xin Gao^*, Yu Li^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

33 Scopus citations

Abstract

We present a novel self-supervised Contrastive LEArning framework for single-cell ribonucleic acid (RNA)-sequencing (CLEAR) data representation and the downstream analysis. Compared with current methods, CLEAR overcomes the heterogeneity of the experimental data with a specifically designed representation learning task and thus can handle batch effects and dropout events simultaneously. It achieves superior performance on a broad range of fundamental tasks, including clustering, visualization, dropout correction, batch effect removal, and pseudo-time inference. The proposed method successfully identifies and illustrates inflammatory-related mechanisms in a COVID-19 disease study with 43 695 single cells from peripheral blood mononuclear cells.

Original language	English (US)
Article number	bbac377
Journal	Briefings in bioinformatics
Volume	23
Issue number	5
DOIs	https://doi.org/10.1093/bib/bbac377
State	Published - Sep 1 2022

Keywords

batch effect removal
contrastive learning
deep learning
scRNA-seq

ASJC Scopus subject areas

Information Systems
Molecular Biology

Access to Document

10.1093/bib/bbac377

Cite this

@article{496ec06ccaf44820bcdb748f9bce2cf3,

title = "Self-supervised contrastive learning for integrative single cell RNA-seq data analysis",

abstract = "We present a novel self-supervised Contrastive LEArning framework for single-cell ribonucleic acid (RNA)-sequencing (CLEAR) data representation and the downstream analysis. Compared with current methods, CLEAR overcomes the heterogeneity of the experimental data with a specifically designed representation learning task and thus can handle batch effects and dropout events simultaneously. It achieves superior performance on a broad range of fundamental tasks, including clustering, visualization, dropout correction, batch effect removal, and pseudo-time inference. The proposed method successfully identifies and illustrates inflammatory-related mechanisms in a COVID-19 disease study with 43 695 single cells from peripheral blood mononuclear cells.",

keywords = "batch effect removal, contrastive learning, deep learning, scRNA-seq",

author = "Wenkai Han and Yuqi Cheng and Jiayang Chen and Huawen Zhong and Zhihang Hu and Siyuan Chen and Licheng Zong and Liang Hong and Chan, {Ting Fung} and Irwin King and Xin Gao and Yu Li",

note = "Publisher Copyright: {\textcopyright} 2022 The Author(s). Published by Oxford University Press.",

year = "2022",

month = sep,

day = "1",

doi = "10.1093/bib/bbac377",

language = "English (US)",

volume = "23",

journal = "Briefings in bioinformatics",

issn = "1467-5463",

publisher = "Oxford University Press",

number = "5",

}

TY - JOUR

T1 - Self-supervised contrastive learning for integrative single cell RNA-seq data analysis

AU - Han, Wenkai

AU - Cheng, Yuqi

AU - Chen, Jiayang

AU - Zhong, Huawen

AU - Hu, Zhihang

AU - Chen, Siyuan

AU - Zong, Licheng

AU - Hong, Liang

AU - Chan, Ting Fung

AU - King, Irwin

AU - Gao, Xin

AU - Li, Yu

PY - 2022/9/1

Y1 - 2022/9/1

N2 - We present a novel self-supervised Contrastive LEArning framework for single-cell ribonucleic acid (RNA)-sequencing (CLEAR) data representation and the downstream analysis. Compared with current methods, CLEAR overcomes the heterogeneity of the experimental data with a specifically designed representation learning task and thus can handle batch effects and dropout events simultaneously. It achieves superior performance on a broad range of fundamental tasks, including clustering, visualization, dropout correction, batch effect removal, and pseudo-time inference. The proposed method successfully identifies and illustrates inflammatory-related mechanisms in a COVID-19 disease study with 43 695 single cells from peripheral blood mononuclear cells.

AB - We present a novel self-supervised Contrastive LEArning framework for single-cell ribonucleic acid (RNA)-sequencing (CLEAR) data representation and the downstream analysis. Compared with current methods, CLEAR overcomes the heterogeneity of the experimental data with a specifically designed representation learning task and thus can handle batch effects and dropout events simultaneously. It achieves superior performance on a broad range of fundamental tasks, including clustering, visualization, dropout correction, batch effect removal, and pseudo-time inference. The proposed method successfully identifies and illustrates inflammatory-related mechanisms in a COVID-19 disease study with 43 695 single cells from peripheral blood mononuclear cells.

KW - batch effect removal

KW - contrastive learning

KW - deep learning

KW - scRNA-seq

UR - http://www.scopus.com/inward/record.url?scp=85138494885&partnerID=8YFLogxK

U2 - 10.1093/bib/bbac377

DO - 10.1093/bib/bbac377

M3 - Article

C2 - 36089561

AN - SCOPUS:85138494885

SN - 1467-5463

VL - 23

JO - Briefings in bioinformatics

JF - Briefings in bioinformatics

IS - 5

M1 - bbac377

ER -

Self-supervised contrastive learning for integrative single cell RNA-seq data analysis

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this