Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning

Xinmin Huang; Xu Cai; Xinpeng Chen; Yiwei Hong; Zhengbo Yan; Jianwei Xiao

doi:doi:10.11648/j.ajbio.20231106.12

Research Article |

| Peer-Reviewed

Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning

Xinmin Huang, Xu Cai, Xinpeng Chen, Yiwei Hong, Zhengbo Yan, Jianwei Xiao^*

Published in American Journal of BioScience (Volume 11, Issue 6)

Received: 31 October 2023 Accepted: 15 November 2023 Published: 24 November 2023

Views: Downloads:

Download PDF

Share This Article

Twitter
Linked In
Facebook

Abstract

Objective: This study aims to identify cuproptosis-related genes in Systemic Sclerosis (SSc) and construct a clinical prediction model. Methods: The GSE33463 dataset was retrieved from the GEO database, and gene set enrichment analysis (GSEA) was used to analyze the expression of pathways related to cuproptosis. Cuproptosis-related genes were extracted, and potential key genes for SSc were selected using the LASSO and Boruta methods to construct a clinical prediction model. The model's predictive ability was evaluated using K-nearest neighbors (KNN) and Lightgbm methods, with assessment based on ROC curves, PR curves, confusion matrices, F-values, and 5-fold cross-validation. The importance of model variables was evaluated using SHAP analysis. Results: Cuproptosis-related pathways were upregulated in SSc. Four key cuproptosis-related genes (PDHB, DLST, PDHA1, DBT) were identified using the LASSO and Boruta methods, leading to the construction of a clinical prediction model through multivariable logistic regression. The model exhibited a C-index of 0.91, an AUC of 0.914 under the ROC curve, and strong performance in 5-fold cross-validation. KNN and Lightgbm models achieved AUC values of 0.9243 and 0.9763, respectively. PR curve AUC values of 0.8492 and 0.9480 demonstrated high precision, while confusion matrix results revealed KNN and Lightgbm model accuracies of 0.8663 and 0.932, respectively. The models provide a basis for the early diagnosis of SSc. Conclusion: The clinical prediction model, based on four cuproptosis-related genes, demonstrates high predictive capability, aiding in the early diagnosis of SSc patients.

Published in	American Journal of BioScience (Volume 11, Issue 6)
DOI	10.11648/j.ajbio.20231106.12
Page(s)	142-149
Creative Commons	This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium or format, provided the original work is properly cited.
Copyright	Copyright © The Author(s), 2024. Published by Science Publishing Group

Keywords

Systemic Sclerosis, Cuproptosis, Machine Learning, Prediction Model, Confusion Matrix

References

[1]	Korman B. Evolving insights into the cellular and molecular pathogenesis of fibrosis in systemic sclerosis. Transl Res. 2019; 209: 77-89. doi: 10.1016/j.trsl.2019.02.010.
[2]	Denton CP, Khanna D. Systemic sclerosis. Lancet. 2017; 390 (10103): 1685-1699. doi: 10.1016/S0140-6736(17)30933-9.
[3]	Bairkdar M, Rossides M, Westerlind H, et al M. Incidence and prevalence of systemic sclerosis globally: a comprehensive systematic review and meta-analysis. Rheumatology (Oxford). 2021; 60 (7): 3121-3133. doi: 10.1093/rheumatology/keab190.
[4]	Pokeerbux MR, Giovannelli J, Dauchet L, et al. Survival and prognosis factors in systemic sclerosis: data of a French multicenter cohort, systematic review, and meta-analysis of the literature. Arthritis Res Ther. 2019; 21 (1): 86. doi: 10.1186/s13075-019-1867-16.
[5]	Tsvetkov P, Coy S, Petrova B, et al. Copper induces cell death by targeting lipoylated TCA cycle proteins [published correction appears in Science. 2022 Apr 22; 376 (6591): eabq4855]. Science. 2022; 375 (6586): 1254-1261. doi: 10.1126/science.abf0529.
[6]	Jiang M, Liu K, Lu S, et al. Verification of cuproptosis-related diagnostic model associated with immune infiltration in rheumatoid arthritis. Front Endocrinol (Lausanne). 2023; 14: 1204926. doi: 10.3389/fendo.2023.1204926.
[7]	Li W, Guan X, Wang Y, et al. Cuproptosis-related gene identification and immune infiltration analysis in systemic lupus erythematosus. Front Immunol. 2023; 14: 1157196. doi: 10.3389/fimmu.2023.1157196.
[8]	Svegliati S, Spadoni T, Moroncini G, Gabrielli A. NADPH oxidase, oxidative stress and fibrosis in systemic sclerosis. Free Radic Biol Med. 2018; 125: 90-97. doi: 10.1016/j.freeradbiomed.2018.04.554.
[9]	Ames PRJ, Bucci T, Merashli M, et al. Oxidative/nitrative stress in the pathogenesis of systemic sclerosis: are antioxidants beneficial?. Free Radic Res. 2018; 52 (10): 1063-1082. doi: 10.1080/10715762.2018.1525712.
[10]	Doridot L, Jeljeli M, Chêne C, Batteux F. Implication of oxidative stress in the pathogenesis of systemic sclerosis via inflammation, autoimmunity and fibrosis. Redox Biol. 2019; 25: 101122. doi: 10.1016/j.redox.2019.101122.
[11]	Abdulle AE, Diercks GFH, Feelisch M, et al. The Role of Oxidative Stress in the Development of Systemic Sclerosis Related Vasculopathy. Front Physiol. 2018; 9: 1177. Published 2018 Aug 24. doi: 10.3389/fphys.2018.01177.
[12]	Li X, Wang X, Li J, et al. Serum Concentrations of Trace Elements/Minerals in Patients with Diffuse Systemic Sclerosis. Biol Trace Elem Res. 2021; 199 (7): 2440-2443. doi: 10.1007/s12011-020-02368-5.
[13]	Zhu Y, Wu G, Yan W, et al. miR-146b-5p regulates cell growth, invasion, and metabolism by targeting PDHB in colorectal cancer. Am J Cancer Res. 2017; 7 (5): 1136-1150. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5446479/
[14]	Saunier E, Benelli C, Bortoli S. The pyruvate dehydrogenase complex in cancer: An old metabolic gatekeeper regulated by new pathways and pharmacological agents. Int J Cancer. 2016 Feb 15; 138 (4): 809-17. doi: 10.1002/ijc.29564.
[15]	Echeverri Ruiz NP, Mohan V, Wu J, et al. Dynamic regulation of mitochondrial pyruvate metabolism is necessary for orthotopic pancreatic tumor growth. Cancer Metab. 2021 Nov 8; 9 (1): 39. doi: 10.1186/s40170-021-00275-4.
[16]	Feng X, Zhang L, Xu S, et al. ATP-citrate lyase (ACLY) in lipid metabolism and atherosclerosis: An updated review. Prog Lipid Res. 2020 Jan; 77: 101006. doi: 10.1016/j.plipres.2019.101006.
[17]	Patel MS, Nemeria NS, Furey W, et al. The pyruvate dehydrogenase complexes: structure-based function and regulation. J Biol Chem. 2014 Jun 13; 289 (24): 16615-23. doi: 10.1074/jbc.R114.563148.
[18]	Infantino V, Iacobazzi V, Menga A, et al. A key role of the mitochondrial citrate carrier (SLC25A1) in TNFα- and IFNγ-triggered inflammation. Biochim Biophys Acta. 2014 Nov; 1839 (11): 1217-1225. doi: 10.1016/j.bbagrm.2014.07.013.
[19]	Iacobazzi V, Infantino V. Citrate--new functions for an old metabolite. Biol Chem. 2014 Apr; 395 (4): 387-99. doi: 10.1515/hsz-2013-0271.
[20]	Ouyang X, Dai Y, Wen JL, et al. ¹H NMR-based metabolomic study of metabolic profiling for systemic lupus erythematosus. Lupus. 2011 Nov; 20 (13): 1411-20. doi: 10.1177/0961203311418707.
[21]	Udumula MP, Sakr S, Dar S, et al. Ovarian cancer modulates the immunosuppressive function of CD11b+Gr1+ myeloid cells via glutamine metabolism. Mol Metab. 2021 Nov; 53: 101272. doi: 10.1016/j.molmet.2021.101272.
[22]	Colliou N, Ge Y, Sahay B, et al. Commensal Propionibacterium strain UF1 mitigates intestinal inflammation via Th17 cell regulation. J Clin Invest. 2017 Nov 1; 127 (11): 3970-3986. doi: 10.1172/JCI95376..
[23]	Ahn SH, Yang HY, Tran GB, et al. Interaction of peroxiredoxin V with dihydrolipoamide branched chain transacylase E2 (DBT) in mouse kidney under hypoxia. Proteome Sci. 2015 Feb 5; 13: 4. doi: 10.1186/s12953-014-0061-2.

Cite This Article

Plain Text BibTeX RIS

APA Style

Huang, X., Cai, X., Chen, X., Hong, Y., Yan, Z., et al. (2023). Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning. American Journal of BioScience, 11(6), 142-149. https://doi.org/10.11648/j.ajbio.20231106.12

Copy | Download

ACS Style

Huang, X.; Cai, X.; Chen, X.; Hong, Y.; Yan, Z., et al. Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning. Am. J. BioScience 2023, 11(6), 142-149. doi: 10.11648/j.ajbio.20231106.12

Copy | Download

AMA Style

Huang X, Cai X, Chen X, Hong Y, Yan Z, et al. Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning. Am J BioScience. 2023;11(6):142-149. doi: 10.11648/j.ajbio.20231106.12

Copy | Download

@article{10.11648/j.ajbio.20231106.12,
  author = {Xinmin Huang and Xu Cai and Xinpeng Chen and Yiwei Hong and Zhengbo Yan and Jianwei Xiao},
  title = {Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning},
  journal = {American Journal of BioScience},
  volume = {11},
  number = {6},
  pages = {142-149},
  doi = {10.11648/j.ajbio.20231106.12},
  url = {https://doi.org/10.11648/j.ajbio.20231106.12},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ajbio.20231106.12},
  abstract = {Objective: This study aims to identify cuproptosis-related genes in Systemic Sclerosis (SSc) and construct a clinical prediction model. Methods: The GSE33463 dataset was retrieved from the GEO database, and gene set enrichment analysis (GSEA) was used to analyze the expression of pathways related to cuproptosis. Cuproptosis-related genes were extracted, and potential key genes for SSc were selected using the LASSO and Boruta methods to construct a clinical prediction model. The model's predictive ability was evaluated using K-nearest neighbors (KNN) and Lightgbm methods, with assessment based on ROC curves, PR curves, confusion matrices, F-values, and 5-fold cross-validation. The importance of model variables was evaluated using SHAP analysis. Results: Cuproptosis-related pathways were upregulated in SSc. Four key cuproptosis-related genes (PDHB, DLST, PDHA1, DBT) were identified using the LASSO and Boruta methods, leading to the construction of a clinical prediction model through multivariable logistic regression. The model exhibited a C-index of 0.91, an AUC of 0.914 under the ROC curve, and strong performance in 5-fold cross-validation. KNN and Lightgbm models achieved AUC values of 0.9243 and 0.9763, respectively. PR curve AUC values of 0.8492 and 0.9480 demonstrated high precision, while confusion matrix results revealed KNN and Lightgbm model accuracies of 0.8663 and 0.932, respectively. The models provide a basis for the early diagnosis of SSc. Conclusion: The clinical prediction model, based on four cuproptosis-related genes, demonstrates high predictive capability, aiding in the early diagnosis of SSc patients.
},
 year = {2023}
}

Copy | Download

TY - JOUR
T1 - Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning
AU - Xinmin Huang
AU - Xu Cai
AU - Xinpeng Chen
AU - Yiwei Hong
AU - Zhengbo Yan
AU - Jianwei Xiao
Y1 - 2023/11/24
PY - 2023
N1 - https://doi.org/10.11648/j.ajbio.20231106.12
DO - 10.11648/j.ajbio.20231106.12
T2 - American Journal of BioScience
JF - American Journal of BioScience
JO - American Journal of BioScience
SP - 142
EP - 149
PB - Science Publishing Group
SN - 2330-0167
UR - https://doi.org/10.11648/j.ajbio.20231106.12
AB - Objective: This study aims to identify cuproptosis-related genes in Systemic Sclerosis (SSc) and construct a clinical prediction model. Methods: The GSE33463 dataset was retrieved from the GEO database, and gene set enrichment analysis (GSEA) was used to analyze the expression of pathways related to cuproptosis. Cuproptosis-related genes were extracted, and potential key genes for SSc were selected using the LASSO and Boruta methods to construct a clinical prediction model. The model's predictive ability was evaluated using K-nearest neighbors (KNN) and Lightgbm methods, with assessment based on ROC curves, PR curves, confusion matrices, F-values, and 5-fold cross-validation. The importance of model variables was evaluated using SHAP analysis. Results: Cuproptosis-related pathways were upregulated in SSc. Four key cuproptosis-related genes (PDHB, DLST, PDHA1, DBT) were identified using the LASSO and Boruta methods, leading to the construction of a clinical prediction model through multivariable logistic regression. The model exhibited a C-index of 0.91, an AUC of 0.914 under the ROC curve, and strong performance in 5-fold cross-validation. KNN and Lightgbm models achieved AUC values of 0.9243 and 0.9763, respectively. PR curve AUC values of 0.8492 and 0.9480 demonstrated high precision, while confusion matrix results revealed KNN and Lightgbm model accuracies of 0.8663 and 0.932, respectively. The models provide a basis for the early diagnosis of SSc. Conclusion: The clinical prediction model, based on four cuproptosis-related genes, demonstrates high predictive capability, aiding in the early diagnosis of SSc patients.

VL - 11
IS - 6
ER -

Copy | Download

Author Information

Xinmin Huang

Rheumatology, Shenzhen Futian Hospital for Rheumatic Diseases, Shenzhen, China
Xu Cai

Rheumatology, Shenzhen Futian Hospital for Rheumatic Diseases, Shenzhen, China
Xinpeng Chen

Rheumatology, Shenzhen Futian Hospital for Rheumatic Diseases, Shenzhen, China
Yiwei Hong

Rheumatology, Shenzhen Futian Hospital for Rheumatic Diseases, Shenzhen, China
Zhengbo Yan

Rheumatology, Shenzhen Futian Hospital for Rheumatic Diseases, Shenzhen, China
Jianwei Xiao

Rheumatology, Shenzhen Futian Hospital for Rheumatic Diseases, Shenzhen, China

Contact Email

Download PDF

Sections

Plain Text BibTeX RIS

APA Style

Huang, X., Cai, X., Chen, X., Hong, Y., Yan, Z., et al. (2023). Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning. American Journal of BioScience, 11(6), 142-149. https://doi.org/10.11648/j.ajbio.20231106.12

Copy | Download

ACS Style

Huang, X.; Cai, X.; Chen, X.; Hong, Y.; Yan, Z., et al. Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning. Am. J. BioScience 2023, 11(6), 142-149. doi: 10.11648/j.ajbio.20231106.12

Copy | Download

AMA Style

Huang X, Cai X, Chen X, Hong Y, Yan Z, et al. Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning. Am J BioScience. 2023;11(6):142-149. doi: 10.11648/j.ajbio.20231106.12

Copy | Download

@article{10.11648/j.ajbio.20231106.12,
  author = {Xinmin Huang and Xu Cai and Xinpeng Chen and Yiwei Hong and Zhengbo Yan and Jianwei Xiao},
  title = {Construction of a Clinical Prediction Model for Systemic Sclerosis Cuproptosis-Related Genes Using Machine Learning},
  journal = {American Journal of BioScience},
  volume = {11},
  number = {6},
  pages = {142-149},
  doi = {10.11648/j.ajbio.20231106.12},
  url = {https://doi.org/10.11648/j.ajbio.20231106.12},
  eprint = {https://article.sciencepublishinggroup.com/pdf/10.11648.j.ajbio.20231106.12},
  abstract = {Objective: This study aims to identify cuproptosis-related genes in Systemic Sclerosis (SSc) and construct a clinical prediction model. Methods: The GSE33463 dataset was retrieved from the GEO database, and gene set enrichment analysis (GSEA) was used to analyze the expression of pathways related to cuproptosis. Cuproptosis-related genes were extracted, and potential key genes for SSc were selected using the LASSO and Boruta methods to construct a clinical prediction model. The model's predictive ability was evaluated using K-nearest neighbors (KNN) and Lightgbm methods, with assessment based on ROC curves, PR curves, confusion matrices, F-values, and 5-fold cross-validation. The importance of model variables was evaluated using SHAP analysis. Results: Cuproptosis-related pathways were upregulated in SSc. Four key cuproptosis-related genes (PDHB, DLST, PDHA1, DBT) were identified using the LASSO and Boruta methods, leading to the construction of a clinical prediction model through multivariable logistic regression. The model exhibited a C-index of 0.91, an AUC of 0.914 under the ROC curve, and strong performance in 5-fold cross-validation. KNN and Lightgbm models achieved AUC values of 0.9243 and 0.9763, respectively. PR curve AUC values of 0.8492 and 0.9480 demonstrated high precision, while confusion matrix results revealed KNN and Lightgbm model accuracies of 0.8663 and 0.932, respectively. The models provide a basis for the early diagnosis of SSc. Conclusion: The clinical prediction model, based on four cuproptosis-related genes, demonstrates high predictive capability, aiding in the early diagnosis of SSc patients.
},
 year = {2023}
}

Copy | Download

VL - 11
IS - 6
ER -

Copy | Download