Comparison of miRNA profiling methods using synthetic miRNA pools and standardized exRNA samples reveals substantial performance differences

Abstract

MicroRNAs (miRNAs) found in biofluids play functional roles in health and in disease pathogenesis, underpinning their potential as clinical biomarkers. Several platforms for measurement of extracellular RNAs have recently become available. We evaluated the reproducibility, accuracy, sensitivity, and specificity of four miRNA quantification platforms, including one widely used discovery approach (small RNA-seq) and three targeted platforms (FirePlex, EdgeSeq, and nCounter). Using pools of synthetic miRNAs, we observed that reproducibility was highest for RNA-seq and EdgeSeq, that all three targeted platforms had lower bias than RNA-seq, and that RNA-seq had the best ability to distinguish between present and absent sequences. Overall reproducibility was lower for plasma samples than synthetic miRNA pools. We compared expression of placental miRNAs in plasma from pregnant and non-pregnant women and observed expected differences with RNA-seq and EdgeSeq, but not FirePlex or nCounter. We conclude that differences in performance among miRNA profiling platforms impact their relative utility as potential assay systems for clinical biomarkers.