Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra
Single-cell RNA-sequencing (scRNA-seq) coupled with robust computational analysis facilitates the characterization of phenotypic heterogeneity within tumors. Current scRNA-seq analysis pipelines are capable of identifying a myriad of malignant and non-malignant cell subtypes from single-cell profili...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
eLife Sciences Publications Ltd
2025-06-01
|
| Series: | eLife |
| Subjects: | |
| Online Access: | https://elifesciences.org/articles/98469 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850209667865640960 |
|---|---|
| author | Namrata Bhattacharya Anja Rockstroh Sanket Suhas Deshpande Sam Koshy Thomas Anunay Yadav Chitrita Goswami Smriti Chawla Pierre Solomon Cynthia Fourgeux Gaurav Ahuja Brett Hollier Himanshu Kumar Antoine Roquilly Jeremie Poschmann Melanie Lehman Colleen C Nelson Debarka Sengupta |
| author_facet | Namrata Bhattacharya Anja Rockstroh Sanket Suhas Deshpande Sam Koshy Thomas Anunay Yadav Chitrita Goswami Smriti Chawla Pierre Solomon Cynthia Fourgeux Gaurav Ahuja Brett Hollier Himanshu Kumar Antoine Roquilly Jeremie Poschmann Melanie Lehman Colleen C Nelson Debarka Sengupta |
| author_sort | Namrata Bhattacharya |
| collection | DOAJ |
| description | Single-cell RNA-sequencing (scRNA-seq) coupled with robust computational analysis facilitates the characterization of phenotypic heterogeneity within tumors. Current scRNA-seq analysis pipelines are capable of identifying a myriad of malignant and non-malignant cell subtypes from single-cell profiling of tumors. However, given the extent of intra-tumoral heterogeneity, it is challenging to assess the risk associated with individual cell subpopulations, primarily due to the complexity of the cancer phenotype space and the lack of clinical annotations associated with tumor scRNA-seq studies. To this end, we introduce SCellBOW, a scRNA-seq analysis framework inspired by document embedding techniques from the domain of Natural Language Processing (NLP). SCellBOW is a novel computational approach that facilitates effective identification and high-quality visualization of single-cell subpopulations. We compared SCellBOW with existing best practice methods for its ability to precisely represent phenotypically divergent cell types across multiple scRNA-seq datasets, including our in-house generated human splenocyte and matched peripheral blood mononuclear cell (PBMC) dataset. For tumor cells, SCellBOW estimates the relative risk associated with each cluster and stratifies them based on their aggressiveness. This is achieved by simulating how the presence or absence of a specific cell subpopulation influences disease prognosis. Using SCellBOW, we identified a hitherto unknown and pervasive AR−/NElow (androgen-receptor-negative, neuroendocrine-low) malignant subpopulation in metastatic prostate cancer with conspicuously high aggressiveness. Overall, the risk-stratification capabilities of SCellBOW hold promise for formulating tailored therapeutic interventions by identifying clinically relevant tumor subpopulations and their impact on prognosis. |
| format | Article |
| id | doaj-art-7c94b57a3fda4c91b2dd46b3b4d972e8 |
| institution | OA Journals |
| issn | 2050-084X |
| language | English |
| publishDate | 2025-06-01 |
| publisher | eLife Sciences Publications Ltd |
| record_format | Article |
| series | eLife |
| spelling | doaj-art-7c94b57a3fda4c91b2dd46b3b4d972e82025-08-20T02:09:58ZengeLife Sciences Publications LtdeLife2050-084X2025-06-011310.7554/eLife.98469Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebraNamrata Bhattacharya0https://orcid.org/0000-0002-5666-2551Anja Rockstroh1Sanket Suhas Deshpande2Sam Koshy Thomas3Anunay Yadav4Chitrita Goswami5Smriti Chawla6Pierre Solomon7Cynthia Fourgeux8Gaurav Ahuja9https://orcid.org/0000-0002-2837-9361Brett Hollier10Himanshu Kumar11https://orcid.org/0000-0001-5246-2694Antoine Roquilly12https://orcid.org/0000-0002-1029-6242Jeremie Poschmann13https://orcid.org/0000-0002-9613-5297Melanie Lehman14Colleen C Nelson15Debarka Sengupta16https://orcid.org/0000-0002-6353-5411Australian Prostate Cancer Research Centre-Queensland, Faculty of Health, School of Biomedical Sciences, Centre for Genomics and Personalised Health, Queensland University of Technology, Brisbane, Australia; Department of Computer Science and Engineering, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, India; Translational Research Institute, Princess Alexandra Hospital, Woolloongabba, AustraliaAustralian Prostate Cancer Research Centre-Queensland, Faculty of Health, School of Biomedical Sciences, Centre for Genomics and Personalised Health, Queensland University of Technology, Brisbane, Australia; Translational Research Institute, Princess Alexandra Hospital, Woolloongabba, AustraliaDepartment of Computational Biology, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, IndiaSchool of Mathematical Sciences, The University of Adelaide, Adelaide, AustraliaDepartment of Computer Science and Engineering, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, IndiaDepartment of Computer Science and Engineering, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, IndiaCenter for Computational Biomedicine, Harvard Medical School, Boston, United StatesNantes Université, CHU Nantes, INSERM, Center for Research in Transplantation and Translational Immunology, UMR, Nantes, FranceNantes Université, CHU Nantes, INSERM, Center for Research in Transplantation and Translational Immunology, UMR, Nantes, FranceDepartment of Computational Biology, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, India; Centre for Artificial Intelligence, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, IndiaAustralian Prostate Cancer Research Centre-Queensland, Faculty of Health, School of Biomedical Sciences, Centre for Genomics and Personalised Health, Queensland University of Technology, Brisbane, Australia; Translational Research Institute, Princess Alexandra Hospital, Woolloongabba, AustraliaLaboratory of Immunology and Infectious Disease Biology, Department of Biological Sciences, Indian Institute of Science Education and Research (IISER), Bhopal, IndiaNantes Université, CHU Nantes, INSERM, Center for Research in Transplantation and Translational Immunology, UMR, Nantes, FranceNantes Université, CHU Nantes, INSERM, Center for Research in Transplantation and Translational Immunology, UMR, Nantes, FranceAustralian Prostate Cancer Research Centre-Queensland, Faculty of Health, School of Biomedical Sciences, Centre for Genomics and Personalised Health, Queensland University of Technology, Brisbane, Australia; Vancouver Prostate Centre, Department of Urologic Sciences, University of British Columbia, Vancouver, CanadaAustralian Prostate Cancer Research Centre-Queensland, Faculty of Health, School of Biomedical Sciences, Centre for Genomics and Personalised Health, Queensland University of Technology, Brisbane, Australia; Translational Research Institute, Princess Alexandra Hospital, Woolloongabba, AustraliaDepartment of Computer Science and Engineering, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, India; Department of Computational Biology, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, India; Centre for Artificial Intelligence, Indraprastha Institute of Information Technology-Delhi (IIIT-Delhi), Okhla, Phase III, New Delhi, IndiaSingle-cell RNA-sequencing (scRNA-seq) coupled with robust computational analysis facilitates the characterization of phenotypic heterogeneity within tumors. Current scRNA-seq analysis pipelines are capable of identifying a myriad of malignant and non-malignant cell subtypes from single-cell profiling of tumors. However, given the extent of intra-tumoral heterogeneity, it is challenging to assess the risk associated with individual cell subpopulations, primarily due to the complexity of the cancer phenotype space and the lack of clinical annotations associated with tumor scRNA-seq studies. To this end, we introduce SCellBOW, a scRNA-seq analysis framework inspired by document embedding techniques from the domain of Natural Language Processing (NLP). SCellBOW is a novel computational approach that facilitates effective identification and high-quality visualization of single-cell subpopulations. We compared SCellBOW with existing best practice methods for its ability to precisely represent phenotypically divergent cell types across multiple scRNA-seq datasets, including our in-house generated human splenocyte and matched peripheral blood mononuclear cell (PBMC) dataset. For tumor cells, SCellBOW estimates the relative risk associated with each cluster and stratifies them based on their aggressiveness. This is achieved by simulating how the presence or absence of a specific cell subpopulation influences disease prognosis. Using SCellBOW, we identified a hitherto unknown and pervasive AR−/NElow (androgen-receptor-negative, neuroendocrine-low) malignant subpopulation in metastatic prostate cancer with conspicuously high aggressiveness. Overall, the risk-stratification capabilities of SCellBOW hold promise for formulating tailored therapeutic interventions by identifying clinically relevant tumor subpopulations and their impact on prognosis.https://elifesciences.org/articles/98469single-cell RNA-seqrisk stratificationtransfer learningprostate cancermarker-freelanguage model |
| spellingShingle | Namrata Bhattacharya Anja Rockstroh Sanket Suhas Deshpande Sam Koshy Thomas Anunay Yadav Chitrita Goswami Smriti Chawla Pierre Solomon Cynthia Fourgeux Gaurav Ahuja Brett Hollier Himanshu Kumar Antoine Roquilly Jeremie Poschmann Melanie Lehman Colleen C Nelson Debarka Sengupta Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra eLife single-cell RNA-seq risk stratification transfer learning prostate cancer marker-free language model |
| title | Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra |
| title_full | Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra |
| title_fullStr | Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra |
| title_full_unstemmed | Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra |
| title_short | Artificial intelligence driven tumor risk stratification from single-cell transcriptomics using phenotype algebra |
| title_sort | artificial intelligence driven tumor risk stratification from single cell transcriptomics using phenotype algebra |
| topic | single-cell RNA-seq risk stratification transfer learning prostate cancer marker-free language model |
| url | https://elifesciences.org/articles/98469 |
| work_keys_str_mv | AT namratabhattacharya artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT anjarockstroh artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT sanketsuhasdeshpande artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT samkoshythomas artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT anunayyadav artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT chitritagoswami artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT smritichawla artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT pierresolomon artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT cynthiafourgeux artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT gauravahuja artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT bretthollier artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT himanshukumar artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT antoineroquilly artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT jeremieposchmann artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT melanielehman artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT colleencnelson artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra AT debarkasengupta artificialintelligencedriventumorriskstratificationfromsinglecelltranscriptomicsusingphenotypealgebra |