AUTHOR=Wang Haotian , Li Shaoshuo , Chen Baixing , Wu Mao , Yin Heng , Shao Yang , Wang Jianwei 

TITLE=Exploring the shared gene signatures of smoking-related osteoporosis and chronic obstructive pulmonary disease using machine learning algorithms

JOURNAL=Frontiers in Molecular Biosciences

VOLUME=Volume 10 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2023.1204031

DOI=10.3389/fmolb.2023.1204031

ISSN=2296-889X

ABSTRACT=Cigarette smoking has been recognized as a predisposing factor for both OP and COPD.The present investigation aimed to elucidate the shared gene signatures affected by cigarette smoking in OP and COPD through gene expression profiling.To this end, one smoking-related OP dataset and one COPD dataset were retrieved from the GEO database. The DEGs were analyzed, and WGCNA was performed. The candidate biomarkers were further identified using the LASSO regression method and a RF machine learning algorithm. The diagnostic value of the method was assessed by constructing a logistic regression model and analyzing the ROC curve. The infiltration of immune cells was finally analyzed for identifying the immune cells that are dysregulated in COPD caused by cigarette smoking. In the smoking-related OP and COPD datasets, respectively, 2858 and 280 DEGs were identified. The results of WGCNA revealed that 982 genes strongly correlated with smoking-related OP. There were 32 DEGs in the COPD dataset that overlapped with the hub genes of OP, and the findings of the GO enrichment analysis indicated that the overlapping genes were mainly enriched in the biological process category of the immune system term. Six candidate genes were identified and used to construct a logistic regression model through LASSO regression and RF machine learning algorithms, and the diagnostic value of the developed model was subsequently determined. The logistic regression model had high diagnostic values with both the training set and external validation datasets, and the AUCs were determined to be 0.83 and 0.99, respectively. The results of the immune cell infiltration analysis indicated that several immune cells exhibited dysregulation. A total of six immune-associated genes, namely, MALT1, PLAT, SCNN1A, SIX3, SPAG9, and VPS35, were identified for smoking-related OP and COPD. The findings revealed that the shared mechanism of pathogenesis of smoking-related OP and COPD is highly influenced by the immune cell infiltration profiles. In addition to providing new insights into the pathogenesis of smoking-related OP and COPD, the results of the study could provide valuable insights into the development of novel therapeutic strategies for the management of these disorders.