Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data.
Purpose: Meta-analyses failed to accurately identify patients with non-metastatic breast cancer who are likely to benefit from chemotherapy, and metabolomics could provide new answers. In our previous published work, patients were clustered using five different unsupervised machine learning (ML) met...
Main Authors: | , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier
2023-01-01
|
Series: | Computational and Structural Biotechnology Journal |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S2001037023003884 |
_version_ | 1797384073041149952 |
---|---|
author | Caroline Bailleux David Chardin Jean-Marie Guigonis Jean-Marc Ferrero Yann Chateau Olivier Humbert Thierry Pourcher Jocelyn Gal |
author_facet | Caroline Bailleux David Chardin Jean-Marie Guigonis Jean-Marc Ferrero Yann Chateau Olivier Humbert Thierry Pourcher Jocelyn Gal |
author_sort | Caroline Bailleux |
collection | DOAJ |
description | Purpose: Meta-analyses failed to accurately identify patients with non-metastatic breast cancer who are likely to benefit from chemotherapy, and metabolomics could provide new answers. In our previous published work, patients were clustered using five different unsupervised machine learning (ML) methods resulting in the identification of three clusters with distinct clinical and simulated survival data. The objective of this study was to evaluate the survival outcomes, with extended follow-up, using the same 5 different methods of unsupervised machine learning. Experimental design: Forty-nine patients, diagnosed between 2013 and 2016, with non-metastatic BC were included retrospectively. Median follow-up was extended to 85.8 months. 449 metabolites were extracted from tumor resection samples by combined Liquid chromatography-mass spectrometry (LC–MS). Survival analyses were reported grouping together Cluster 1 and 2 versus cluster 3. Bootstrap optimization was applied. Results: PCA k-means, K-sparse and Spectral clustering were the most effective methods to predict 2-year progression-free survival with bootstrap optimization (PFSb); as bootstrap example, with PCA k-means method, PFSb were 94% for cluster 1&2 versus 82% for cluster 3 (p = 0.01). PCA k-means method performed best, with higher reproducibility (mean HR=2 (95%CI [1.4–2.7]); probability of p ≤ 0.05 85%). Cancer-specific survival (CSS) and overall survival (OS) analyses highlighted a discrepancy between the 5 ML unsupervised methods. Conclusion: Our study is a proof-of-principle that it is possible to use unsupervised ML methods on metabolomic data to predict PFS survival outcomes, with the best performance for PCA k-means. A larger population study is needed to draw conclusions from CSS and OS analyses. |
first_indexed | 2024-03-08T21:30:13Z |
format | Article |
id | doaj.art-816f939c016a42129efa98d1890845fc |
institution | Directory Open Access Journal |
issn | 2001-0370 |
language | English |
last_indexed | 2024-03-08T21:30:13Z |
publishDate | 2023-01-01 |
publisher | Elsevier |
record_format | Article |
series | Computational and Structural Biotechnology Journal |
spelling | doaj.art-816f939c016a42129efa98d1890845fc2023-12-21T07:32:22ZengElsevierComputational and Structural Biotechnology Journal2001-03702023-01-012151365143Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data.Caroline Bailleux0David Chardin1Jean-Marie Guigonis2Jean-Marc Ferrero3Yann Chateau4Olivier Humbert5Thierry Pourcher6Jocelyn Gal7University Côte d′Azur, Centre Antoine Lacassagne, Medical Oncology Department, Nice F-06189, France; University Côte d′Azur, Commissariat à l′Energie Atomique et aux énergies alternatives, Institut Frédéric Joliot, Service Hospitalier Frédéric Joliot, laboratory Transporters in Oncology and Radiotherapy in Oncology (TIRO), School of medicine, Nice F-06100, France; Correspondence to: Medical Oncology Department, Centre Antoine Lacassagne, University Côte d′Azur, 33 avenue de Valombrose, 06189 Nice, France.University Côte d′Azur, Centre Antoine Lacassagne, Nuclear medicine Department, Nice F-06189, France; University Côte d′Azur, Commissariat à l′Energie Atomique et aux énergies alternatives, Institut Frédéric Joliot, Service Hospitalier Frédéric Joliot, laboratory Transporters in Oncology and Radiotherapy in Oncology (TIRO), School of medicine, Nice F-06100, FranceUniversity Côte d′Azur, Commissariat à l′Energie Atomique et aux énergies alternatives, Institut Frédéric Joliot, Service Hospitalier Frédéric Joliot, laboratory Transporters in Oncology and Radiotherapy in Oncology (TIRO), School of medicine, Nice F-06100, FranceUniversity Côte d′Azur, Centre Antoine Lacassagne, Medical Oncology Department, Nice F-06189, FranceUniversity Côte d′Azur, Centre Antoine Lacassagne, Epidemiology and Biostatistics Department, Nice F-06189, FranceUniversity Côte d′Azur, Centre Antoine Lacassagne, Nuclear medicine Department, Nice F-06189, France; University Côte d′Azur, Commissariat à l′Energie Atomique et aux énergies alternatives, Institut Frédéric Joliot, Service Hospitalier Frédéric Joliot, laboratory Transporters in Oncology and Radiotherapy in Oncology (TIRO), School of medicine, Nice F-06100, FranceUniversity Côte d′Azur, Commissariat à l′Energie Atomique et aux énergies alternatives, Institut Frédéric Joliot, Service Hospitalier Frédéric Joliot, laboratory Transporters in Oncology and Radiotherapy in Oncology (TIRO), School of medicine, Nice F-06100, FranceUniversity Côte d′Azur, Centre Antoine Lacassagne, Epidemiology and Biostatistics Department, Nice F-06189, FrancePurpose: Meta-analyses failed to accurately identify patients with non-metastatic breast cancer who are likely to benefit from chemotherapy, and metabolomics could provide new answers. In our previous published work, patients were clustered using five different unsupervised machine learning (ML) methods resulting in the identification of three clusters with distinct clinical and simulated survival data. The objective of this study was to evaluate the survival outcomes, with extended follow-up, using the same 5 different methods of unsupervised machine learning. Experimental design: Forty-nine patients, diagnosed between 2013 and 2016, with non-metastatic BC were included retrospectively. Median follow-up was extended to 85.8 months. 449 metabolites were extracted from tumor resection samples by combined Liquid chromatography-mass spectrometry (LC–MS). Survival analyses were reported grouping together Cluster 1 and 2 versus cluster 3. Bootstrap optimization was applied. Results: PCA k-means, K-sparse and Spectral clustering were the most effective methods to predict 2-year progression-free survival with bootstrap optimization (PFSb); as bootstrap example, with PCA k-means method, PFSb were 94% for cluster 1&2 versus 82% for cluster 3 (p = 0.01). PCA k-means method performed best, with higher reproducibility (mean HR=2 (95%CI [1.4–2.7]); probability of p ≤ 0.05 85%). Cancer-specific survival (CSS) and overall survival (OS) analyses highlighted a discrepancy between the 5 ML unsupervised methods. Conclusion: Our study is a proof-of-principle that it is possible to use unsupervised ML methods on metabolomic data to predict PFS survival outcomes, with the best performance for PCA k-means. A larger population study is needed to draw conclusions from CSS and OS analyses.http://www.sciencedirect.com/science/article/pii/S2001037023003884Unsupervised machine learningClusteringBreast cancerSurvivalUntargetedProof-of-concept |
spellingShingle | Caroline Bailleux David Chardin Jean-Marie Guigonis Jean-Marc Ferrero Yann Chateau Olivier Humbert Thierry Pourcher Jocelyn Gal Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data. Computational and Structural Biotechnology Journal Unsupervised machine learning Clustering Breast cancer Survival Untargeted Proof-of-concept |
title | Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data. |
title_full | Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data. |
title_fullStr | Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data. |
title_full_unstemmed | Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data. |
title_short | Survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data. |
title_sort | survival analysis of patient groups defined by unsupervised machine learning clustering methods based on patient metabolomic data |
topic | Unsupervised machine learning Clustering Breast cancer Survival Untargeted Proof-of-concept |
url | http://www.sciencedirect.com/science/article/pii/S2001037023003884 |
work_keys_str_mv | AT carolinebailleux survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT davidchardin survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT jeanmarieguigonis survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT jeanmarcferrero survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT yannchateau survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT olivierhumbert survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT thierrypourcher survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata AT jocelyngal survivalanalysisofpatientgroupsdefinedbyunsupervisedmachinelearningclusteringmethodsbasedonpatientmetabolomicdata |