“…The sensitivity of mNGS compared to targeted qPCR was low, at 30% overall, and even lower for bacteria at 5%. This was not surprising to us and can likely explained by several factors, including the read depth, limited sample volume, age of the samples (10+ years), different sample types (sequencing on plasma versus qPCR on whole blood), stringency of the background filtering thresholds used, and the potential presence of inhibitory substances to sequencing in human plasma, which have all been previously described ( 21 – 23 ). Several studies have shown similarly variable sensitivity and specificity of mNGS versus PCR ( 24 – 27 ).…”