Laboratory validation of an RNA/DNA hybrid tagmentation based mNGS workflow on SARS-CoV-2 and other respiratory RNA viruses detection

BackgroundAcute respiratory infection (ARI) caused by RNA viruses is still one of the main diseases all over the world such as SARS-CoV-2 and Influenza A virus. mNGS was a powerful tool for ethological diagnosis. But there were some challenges during mNGS implementation in clinical settings such as time-consuming manipulation and lack of comprehensive analytical validation. MethodsWe set up CATCH that was a mNGS method based on RNA and DNA hybrid tagmentation via Tn5 transposon. Seven respiratory RNA viruses and three subtypes of Influenza A virus had been used to test CATCHs capabilities of detection and semiquantification. Analytical performance of SARS-CoV-2 and Influenza A virus had been determined with reference standards. We compared accuracy of CATCH with quantitative real time PCR by using clinical 98 samples from 64 COVID-19 patients. ResultsWe minimized the library preparation process to 3 hours and handling time to 35 minutes. Duplicate filtered RPM of 7 respiratory viruses and 3 Influenza A virus subtypes were highly correlated with viral concentration (r=0.60, p<0.001, Spearman correlation test). LOD of SARS-CoV-2 was 39.2 copies/test and of Influenza A virus was 278.1 copies/mL Comparing with qR-TPCR, the overall accuracy of CATCH was 91.4%. Sensitivity was 84.5% and specificity was 100%. Meanwhile, there were significant difference of microbial profile in oropharyngeal swabs among critical, moderate patients and healthy controls (p<0.001, PERMANOVA test). ConclusionAlthough further optimization is needed before CATCH can be rolled out as a routine diagnostic test, we highlight the potential impact of it advancing molecular diagnostics for respiratory pathogens.