Blood circRNAs as biomarkers for the diagnosis of community‐acquired pneumonia

Abstract Circular RNAs (circRNAs) have been reported as effective diagnostic and therapeutic biomarkers in many diseases, but the potential of using this easy‐to‐monitor and highly stable materials for diagnosing Community‐acquired pneumonia (CAP) remains unexplored. Here, aiming to identify potential CAP‐related circRNAs in peripheral blood and seeking to deepen the understanding of how circRNA‐miRNA‐mRNA regulatory networks may contribute to CAP, we applied microarrays profiling analysis and identified 8296 differentially expressed (DE) circRNAs between patients with CAP (n = 6) and healthy controls (n = 6). Subsequently, we validated the accumulation trends for the top 100 DE circRNAs based on qPCR in an independent validation cohort (30 patients vs 30 controls), and ultimately identified a panel of four circRNAs that perform extremely well as sensitive and specific biomarkers for diagnosing CAP: hsa_circ_0018429 (area under the curve [AUC] = 0.8216), hsa_circ_0026579 (AUC = 0.7733), hsa_circ_0125357 (AUC = 0.7730), and hsa_circ_0099188 (AUC = 0.6978); combined as a panel (AUC = 0.8776). In addition, hsa_circ_0026579 exhibited good performance in differentiating viral from bacterial or mixed infection, with an AUC of 0.863. We also identified 10 miRNAs that most likely to interact with these four circRNAs, and then predicted 205 mRNA target genes. The KEGG pathway enrichment analysis suggested highly plausible functional implications related to inflammation and to virus‐infection‐related signaling pathways (such as HTLV‐1 infection and hepatitis B infection). Thus, we generated a genetic network of potential CAP‐related regulatory interactions that should inform future hypothesis‐driven research into the causes and potential treatment of this widespread and frequently fatal disease.