Immune-Associated Gene Signatures and Subtypes to Predict the Progression of Atherosclerotic Plaques Based on Machine Learning

Yujia Yang; Xu Yi; Yue Cai; Yuan Zhang; Zhiqiang Xu

doi:10.3389/fphar.2022.865624

Immune-Associated Gene Signatures and Subtypes to Predict the Progression of Atherosclerotic Plaques Based on Machine Learning

Front Pharmacol. 2022 Apr 26:13:865624. doi: 10.3389/fphar.2022.865624. eCollection 2022.

Authors

Yujia Yang¹, Xu Yi¹, Yue Cai², Yuan Zhang¹, Zhiqiang Xu¹

Affiliations

¹ Department of Neurology and Centre for Clinical Neuroscience, Daping Hospital, Army Medical University (Third Military Medical University), Chongqing, China.
² Department of Cardiology, Xijing Hospital, Fourth Military Medical University, Xi'an, China.

Abstract

Objective: Experimental and clinical evidence suggests that atherosclerosis is a chronic inflammatory disease. Our study was conducted for uncovering the roles of immune-associated genes during atherosclerotic plaque progression. Methods: Gene expression profiling of GSE28829, GSE43292, GSE41571, and GSE120521 datasets was retrieved from the GEO database. Three machine learning algorithms, least absolute shrinkage, and selection operator (LASSO), random forest, and support vector machine-recursive feature elimination (SVM-RFE) were utilized for screening characteristic genes among atherosclerotic plaque progression- and immune-associated genes. ROC curves were generated for estimating the diagnostic efficacy. Immune cell infiltrations were estimated via ssGSEA, and immune checkpoints were quantified. CMap analysis was implemented to screen potential small-molecule compounds. Atherosclerotic plaque specimens were classified using a consensus clustering approach. Results: Seven characteristic genes (TNFSF13B, CCL5, CCL19, ITGAL, CD14, GZMB, and BTK) were identified, which enabled the prediction of progression of atherosclerotic plaques. Higher immune cell infiltrations and immune checkpoint expressions were found in advanced-stage than in early-stage atherosclerotic plaques and were positively linked to characteristic genes. Patients could clinically benefit from the characteristic gene-based nomogram. Several small molecular compounds were predicted based on the characteristic genes. Two subtypes, namely, C1 immune subtype and C2 non-immune subtype, were classified across atherosclerotic plaques. The characteristic genes presented higher expression in C1 than in C2 subtypes. Conclusion: Our findings provide several promising atherosclerotic plaque progression- and immune-associated genes as well as immune subtypes, which might enable to assist the design of more accurately tailored cardiovascular immunotherapy.

Keywords: atherosclerosis; atherosclerotic plaque; characteristic genes; immune subtype; immune-associated genes.