Comparison of Reproducibility, Accuracy, Sensitivity, and Specificity of miRNA Quantification Platforms

Abstract

Summary: Given the increasing interest in their use as disease biomarkers, the establishment of reproducible, accurate, sensitive, and specific platforms for microRNA (miRNA) quantification in biofluids is of high priority. We compare four platforms for these characteristics: small RNA sequencing (RNA-seq), FirePlex, EdgeSeq, and nCounter. For a pool of synthetic miRNAs, coefficients of variation for technical replicates are lower for EdgeSeq (6.9%) and RNA-seq (8.2%) than for FirePlex (22.4%); nCounter replicates are not performed. Receiver operating characteristic analysis for distinguishing present versus absent miRNAs shows small RNA-seq (area under curve 0.99) is superior to EdgeSeq (0.97), nCounter (0.94), and FirePlex (0.81). Expected differences in expression of placenta-associated miRNAs in plasma from pregnant and non-pregnant women are observed with RNA-seq and EdgeSeq, but not FirePlex or nCounter. These results indicate that differences in performance among miRNA profiling platforms impact ability to detect biological differences among samples and thus their relative utility for research and clinical use. : Using pools of synthetic RNA oligonucleotides and standardized extracellular RNA samples, Godoy et al. compare small RNA sequencing to three targeted miRNA quantification platforms to evaluate reproducibility, bias, specificity and sensitivity, and accuracy. Each platform has strengths and limitations important to consider for biomarker discovery, clinical validation, and broad clinical use. Keywords: small RNA-sequencing, HTG Molecular EdgeSeq, Abcam FirePlex, NanoString nCounter, miRNA quantification, extracellular RNA