The final, formatted version of the article will be published soon.
ORIGINAL RESEARCH article
Front. Immunol.
Sec. Cancer Immunity and Immunotherapy
Volume 15 - 2024 |
doi: 10.3389/fimmu.2024.1472354
This article is part of the Research Topic Studying the immune microenvironment of liver cancer using artificial intelligence View all 4 articles
Screening of genes co-associated with osteoporosis and chronic HBV infection based on bioinformatics analysis and machine learning
Provisionally accepted- 1 First Department of Hepatology, Tianjin Second People's Hospital, Tianjin, Hebei, China
- 2 1. Clinical College of Neurology, Neurosurgery and Neurorehabilitation, Tianjin Medical University, Tianjin, China
Objective: To identify HBV-related genes (HRGs) implicated in osteoporosis (OP) pathogenesis and develop a diagnostic model for early OP detection in chronic HBV infection (CBI) patients.Methods: Five public sequencing datasets were collected from the GEO database. Gene differential expression and LASSO analyses identified genes linked to OP and CBI. Machine learning algorithms (random forests, support vector machines, and gradient boosting machines) further filtered these genes. The best diagnostic model was chosen based on accuracy and Kappa values. A nomogram model based on HRGs was constructed and assessed for reliability. OP patients were divided into two chronic HBV-related clusters using non-negative matrix factorization. Differential gene expression analysis, Gene Ontology, and KEGG enrichment analyses explored the roles of these genes in OP progression, using ssGSEA and GSVA. Differences in immune cell infiltration between clusters and the correlation between HRGs and immune cells were examined using ssGSEA and the Pearson method.Results: Differential gene expression analysis of CBI and combined OP dataset identified 822 and 776 differentially expressed genes, respectively, with 43 genes intersecting. Following LASSO analysis and various machine learning recursive feature elimination algorithms, 16 HRGs were identified. The support vector machine emerged as the best predictive model based on accuracy and Kappa values, with AUC values of 0.92, 0.83, 0.74, and 0.7 for the training set, validation set, GSE7429, and GSE7158, respectively. The nomogram model exhibited AUC values of 0.91, 0.79, and 0.68 in the training set, GSE7429, and GSE7158, respectively. Non-negative matrix factorization divided OP patients into two clusters, revealing statistically significant differences in 11 types of immune cell infiltration between clusters. Finally, intersecting the HRGs obtained from LASSO analysis with the HRGs identified three genes. Conclusion: This study successfully identified HRGs and developed an efficient diagnostic model based on HRGs, demonstrating high accuracy and strong predictive performance across multiple datasets. This research not only offers new insights into the complex relationship between OP and CBI but also establishes a foundation for the development of early diagnostic and personalized treatment strategies for chronic HBV-related OP.
Keywords: Osteoporosis, HBV, bioinformatics, machine learning, Disease typing, Immune Cell Infiltration
Received: 29 Jul 2024; Accepted: 28 Aug 2024.
Copyright: © 2024 Jiang, Yang, Yang, Hu, Tong, Liu, Liu and Sun. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
* Correspondence:
Beibei Jiang, First Department of Hepatology, Tianjin Second People's Hospital, Tianjin, 300192, Hebei, China
Disclaimer: All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.