Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets

Jia Zhao; Gefei Wang; Jingsi Ming; Zhixiang Lin; Yang Wang; The Tabula Microcebus Consortium; Angela Ruohao Wu; Can Yang

doi:10.1038/s43588-022-00251-y

Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets

Jia Zhao, Gefei Wang, Jingsi Ming, Zhixiang Lin, Yang Wang, The Tabula Microcebus Consortium, Angela Ruohao Wu^*, Can Yang^*

^*Corresponding author for this work

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaper › Journal article › Research › peer-review

13 Citations (Scopus)

Abstract

The rapid emergence of large-scale atlas-level single-cell RNA-seq datasets presents remarkable opportunities for broad and deep biological investigations through integrative analyses. However, harmonizing such datasets requires integration approaches to be not only computationally scalable, but also capable of preserving a wide range of fine-grained cell populations. We have created Portal, a unified framework of adversarial domain translation to learn harmonized representations of datasets. When compared to other state-of-the-art methods, Portal achieves better performance for preserving biological variation during integration, while achieving the integration of millions of cells, in minutes, with low memory consumption. We show that Portal is widely applicable to integrating datasets across different samples, platforms and data types. We also apply Portal to the integration of cross-species datasets with limited shared information among them, elucidating biological insights into the similarities and divergences in the spermatogenesis process among mouse, macaque and human.

Original language	English
Journal	Nature Computational Science
Volume	2
Issue	5
Pages (from-to)	317-330
Number of pages	14
DOIs	https://doi.org/10.1038/s43588-022-00251-y
Publication status	Published - May 2022

Access to Document

10.1038/s43588-022-00251-y

https://www.biorxiv.org/content/biorxiv/early/2022/03/20/2021.11.16.468892.full.pdfLicence: CC BY-NC-ND

Cite this

@article{f6f2a502ca6c4b90876f6c597992dcf2,

title = "Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets",

abstract = "The rapid emergence of large-scale atlas-level single-cell RNA-seq datasets presents remarkable opportunities for broad and deep biological investigations through integrative analyses. However, harmonizing such datasets requires integration approaches to be not only computationally scalable, but also capable of preserving a wide range of fine-grained cell populations. We have created Portal, a unified framework of adversarial domain translation to learn harmonized representations of datasets. When compared to other state-of-the-art methods, Portal achieves better performance for preserving biological variation during integration, while achieving the integration of millions of cells, in minutes, with low memory consumption. We show that Portal is widely applicable to integrating datasets across different samples, platforms and data types. We also apply Portal to the integration of cross-species datasets with limited shared information among them, elucidating biological insights into the similarities and divergences in the spermatogenesis process among mouse, macaque and human.",

author = "Jia Zhao and Gefei Wang and Jingsi Ming and Zhixiang Lin and Yang Wang and Snigdha Agarwal and Aditi Agrawal and Ahmad Al-Moujahed and Alina Alam and Albertelli, {Megan A.} and Paul Allegakoen and Thomas Ambrosi and Jane Antony and Steven Artandi and Fabienne Aujard and Kyle Awayan and Ankit Baghel and Isaac Bakerman and Bakken, {Trygve E.} and Jalal Baruni and Philip Beachy and Biter Bilen and Olga Botvinnik and Boyd, {Scott D.} and Deviana Burhan and Casey, {Kerriann M.} and Charles Chan and Chang, {Charles A.} and Stephen Chang and Ming Chen and Clarke, {Michael F.} and Sheela Crasta and Rebecca Culver and Jessica D{\textquoteright}Addabbo and Spyros Darmanis and Roozbeh Dehghannasiri and Ding, {Song Lin} and Duffy, {Connor V.} and Jacques Epelbaum and Espinoza, {F. Hern{\'a}n} and Camille Ezran and Jean Farup and Ferrell, {James E.} and Frank, {Hannah K.} and Margaret Fuller and Astrid Gillich and Elias Godoy and {de Morree}, Antoine and Rando, {Thomas A.} and Bo Wang and {The Tabula Microcebus Consortium} and Wu, {Angela Ruohao} and Can Yang",

year = "2022",

month = may,

doi = "10.1038/s43588-022-00251-y",

language = "English",

volume = "2",

pages = "317--330",

journal = "Nature Computational Science",

issn = "2662-8457",

publisher = "Springer Nature",

number = "5",

}

TY - JOUR

T1 - Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets

AU - Zhao, Jia

AU - Wang, Gefei

AU - Ming, Jingsi

AU - Lin, Zhixiang

AU - Wang, Yang

AU - Agarwal, Snigdha

AU - Agrawal, Aditi

AU - Al-Moujahed, Ahmad

AU - Alam, Alina

AU - Albertelli, Megan A.

AU - Allegakoen, Paul

AU - Ambrosi, Thomas

AU - Antony, Jane

AU - Artandi, Steven

AU - Aujard, Fabienne

AU - Awayan, Kyle

AU - Baghel, Ankit

AU - Bakerman, Isaac

AU - Bakken, Trygve E.

AU - Baruni, Jalal

AU - Beachy, Philip

AU - Bilen, Biter

AU - Botvinnik, Olga

AU - Boyd, Scott D.

AU - Burhan, Deviana

AU - Casey, Kerriann M.

AU - Chan, Charles

AU - Chang, Charles A.

AU - Chang, Stephen

AU - Chen, Ming

AU - Clarke, Michael F.

AU - Crasta, Sheela

AU - Culver, Rebecca

AU - D’Addabbo, Jessica

AU - Darmanis, Spyros

AU - Dehghannasiri, Roozbeh

AU - Ding, Song Lin

AU - Duffy, Connor V.

AU - Epelbaum, Jacques

AU - Espinoza, F. Hernán

AU - Ezran, Camille

AU - Farup, Jean

AU - Ferrell, James E.

AU - Frank, Hannah K.

AU - Fuller, Margaret

AU - Gillich, Astrid

AU - Godoy, Elias

AU - de Morree, Antoine

AU - Rando, Thomas A.

AU - Wang, Bo

AU - The Tabula Microcebus Consortium

AU - Wu, Angela Ruohao

AU - Yang, Can

PY - 2022/5

Y1 - 2022/5

N2 - The rapid emergence of large-scale atlas-level single-cell RNA-seq datasets presents remarkable opportunities for broad and deep biological investigations through integrative analyses. However, harmonizing such datasets requires integration approaches to be not only computationally scalable, but also capable of preserving a wide range of fine-grained cell populations. We have created Portal, a unified framework of adversarial domain translation to learn harmonized representations of datasets. When compared to other state-of-the-art methods, Portal achieves better performance for preserving biological variation during integration, while achieving the integration of millions of cells, in minutes, with low memory consumption. We show that Portal is widely applicable to integrating datasets across different samples, platforms and data types. We also apply Portal to the integration of cross-species datasets with limited shared information among them, elucidating biological insights into the similarities and divergences in the spermatogenesis process among mouse, macaque and human.

AB - The rapid emergence of large-scale atlas-level single-cell RNA-seq datasets presents remarkable opportunities for broad and deep biological investigations through integrative analyses. However, harmonizing such datasets requires integration approaches to be not only computationally scalable, but also capable of preserving a wide range of fine-grained cell populations. We have created Portal, a unified framework of adversarial domain translation to learn harmonized representations of datasets. When compared to other state-of-the-art methods, Portal achieves better performance for preserving biological variation during integration, while achieving the integration of millions of cells, in minutes, with low memory consumption. We show that Portal is widely applicable to integrating datasets across different samples, platforms and data types. We also apply Portal to the integration of cross-species datasets with limited shared information among them, elucidating biological insights into the similarities and divergences in the spermatogenesis process among mouse, macaque and human.

UR - http://www.scopus.com/inward/record.url?scp=85134019135&partnerID=8YFLogxK

U2 - 10.1038/s43588-022-00251-y

DO - 10.1038/s43588-022-00251-y

M3 - Journal article

C2 - 38177826

AN - SCOPUS:85134019135

SN - 2662-8457

VL - 2

SP - 317

EP - 330

JO - Nature Computational Science

JF - Nature Computational Science

IS - 5

ER -

Adversarial domain translation networks for integrating large-scale atlas-level single-cell datasets

Abstract

Access to Document

Other files and links

Fingerprint

Cite this