Single-Cell Omics Analysis Reveals Immunological Dysregulation in COVID-19 and HIV: Identifying a Shared Abnormality of B Cell Activation via the Unfolded Protein Response and Diagnostic Biomarkers Using Machine Learning Algorithms

Abstract

An increasing number of studies have demonstrated the exacerbation of disease progression of both COVID-19 and HIV when coexisting. However, the molecular mechanisms underlying the interplay between these two viruses are still poorly understood. In this study, we utilized a comprehensive analysis of single-cell transcriptomics to identify and characterize peripheral blood cell subsets in COVID-19 and HIV-infected patients. Our findings revealed that COVID-19 and HIV exhibit a partially similar cellular composition and cell cycle distribution. Additionally, we identified a common pathogenesis in B-cell subsets of COVID-19 and HIV patients, which showed abnormal activation states of the unfolded protein response (UPR) pathway. Based on B-cell signature genes and UPR-related genes, we developed a machine learning diagnostic model that can accurately diagnose both COVID-19 and HIV infections. Our model was validated using a large number of bulk transcriptome data sets and showed good clinical efficacy. Our study provides molecular insights into the single-cell level interplay between SARS-CoV-2 and HIV infections, suggesting a possible common disease mechanism that warrants further investigation.