On the Reproducibility of MiRNA-Seq Differential Expression Analyses in Neuropsychiatric Diseases

MiRNAs are attracting considerable interest as potential biomarkers on neuropsychiatric diseases due to their expression plasticity. In the last decade, a large number of studies have been published in this regard with promising results; however, there is widespread concern about the reproducibility of these results. This study aims to compare the differentially expressed miRNAs reported by 5 recent studies of neuropsychiatric diseases, with those obtained through the miARma-Seq pipeline [1]. In general, we found a low reproducibility (0-74%), and some variations related to the software used for the differential expression analysis. Our results support the idea that miRNAs reported as potential biomarkers in neuropsychiatric diseases are strongly correlated with the analytical methodology and the biological references used; nonetheless, further research is needed to establish the magnitude of this problem and spot its main causes.