MicroRNAs have long been regarded as promising biomarkers for disease detection due to their widespread presence in biofluids. However, the field has been hindered by the high levels of redundancy in microRNA regulation across disorders and difficulties in normalising their expression, which have thus far limited their potential. To overcome these limitations, we propose expanding our search beyond the approximately 2,000 known microRNAs in humans by considering disease-specific microRNA length variations.
In this study, we analysed small-RNA sequencing data from publicly available datasets of influenza A and COVID-19 infections. Using bioinformatic analyses and supervised machine learning techniques, we evaluated the ability of pairs of microRNA isoforms to differentiate between samples based on disease state and severity. Our findings demonstrate that a single isoform pair can accurately distinguish influenza A-infected monocytes with over 90% accuracy. Furthermore, a distinct single isoform pair can predict COVID-19 infection status in patients with over 80% accuracy, across samples from independent datasets. Critically, our use of pairs of isoforms derived from the same microRNA locus offers a new approach to microRNA normalisation relying on post-transcriptional regulation of microRNA processing, rather than transcriptional variation. Our data suggest that this novel approach offers greater resilience to differences in baseline microRNA expression and provides a potential solution to the normalisation issues that have hindered the development of microRNA biomarkers to date.