Integrated Bioinformatics and Machine Learning for Ascertainment and Validation of Biomarkers for Screening Breast Disease
Genes · 2025
DOI: 10.3390/genes16111389 ↗Kurz gefasst
Fachartikel in der Zeitschrift Genes, 2025. Verfasst von Qi Wang und 5 weiteren Autor:innen. Beteiligte Einrichtung: National Key Laboratory, Heilongjiang University of Chinese Medicine, Harbin 150000, China. Hauptthemen laut MEDLINE: Machine Learning, Biomarkers, Tumor, Breast Neoplasms, Computational Biology. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 41 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Natural Science Foundation of Heilongjiang Province.
Abstract (Originalsprache)
Background
This research sought to screen potential biomarkers in diagnosing breast diseases and elucidating their immune-related mechanisms.
Methods
Three datasets were attained from the Gene Expression Omnibus (GEO) database. LIMMA package and weighted gene co-expression network analysis (WGCNA) were used to ascertain differentially expressed genes (DEGs) and key modules in benign breast disease (BBD) and breast cancer (BC). The intersecting genes underwent functional enrichment analysis. Three machine learning (ML) methods (encompassing LASSO regression, random forest, and support vector machine recursive feature elimination (SVM-RFE)) were implemented to select core genes. The diagnostic performance of the core genes was evaluated by comparing their expression levels, plotting receiver operating characteristic (ROC) curves, and constructing a Nomogram. The TCGA-BRCA dataset was used to estimate the prognostic capability of the core genes among individuals with BC. Finally, the IC infiltration was ascertained utilizing the CIBERSORT algorithm.
Results
In total, 2579 DEGs were identified in BBD. WGCNA exhibited that the 1652 genes in green and pink modules were strongly correlated with BBD. In BC, 2742 DEGs were identified. The turquoise and red modules contained 7286 genes exhibiting strong correlations with BC. After intersecting, 41 common genes were obtained, which were predominantly enriched in immune and inflammation regulation pathways. Through integrated screening with three ML algorithms, Arrestin Domain Containing 1 (ARRDC1) and ATPase Sarcoplasmic/Endoplasmic Reticulum Ca2+ Transporting 2 (ATP2A2) were identified as core genes. The ROC curve exhibited that the AUC for the two genes was greater than 0.8. The calibration curve of the nomogram signified a strong alignment between the anticipated risk and detected results. Survival analysis in TCGA-BRCA showed that the high expression of the two genes exhibited a significantly positive association with unfavorable prognosis. Immune infiltration analysis further demonstrated the dysregulation of multiple immune cells in patient samples. Conclusions:ARRDC1 and ATP2A2 are strongly linked to BBD and BC. These findings might enhance our comprehension of the pathogenesis and progression of both BBD and BC, offering prospective biological biomarkers and therapeutic targets for clinical treatment.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- National Key Laboratory, Heilongjiang University of Chinese Medicine, Harbin 150000, China.
- State Key Laboratory of Multimodal Artificial Intelligence, Institute of Automation, Chinese Academy of Sciences, Beijing 100000, China.
ORCID-Profile: Qi Wang
Förderung
- Natural Science Foundation of Heilongjiang Province
Zitieren
APA
Wang, Q., Yang, S., Zhang, Y., Piao, C., Liu, X., & Wu, X. (2025). Integrated Bioinformatics and Machine Learning for Ascertainment and Validation of Biomarkers for Screening Breast Disease. Genes. https://doi.org/10.3390/genes16111389
Vancouver
Wang Q, Yang S, Zhang Y, Piao C, Liu X, Wu X. Integrated Bioinformatics and Machine Learning for Ascertainment and Validation of Biomarkers for Screening Breast Disease. Genes. 2025. doi:10.3390/genes16111389
BibTeX
@article{Wang2025,
author = {Wang, Qi and Yang, Saisai and Zhang, Yao and Piao, Chengyu and Liu, Xin and Wu, Xiuhong},
title = {Integrated Bioinformatics and Machine Learning for Ascertainment and Validation of Biomarkers for Screening Breast Disease},
journal = {Genes},
year = {2025},
doi = {10.3390/genes16111389}
}
EUVIMED