# Translational MicroRNA research in COVID-19: bridging in Silico prediction with clinical biomarker potential

## TL;DR

This study explores the potential of specific microRNAs as biomarkers for diagnosing and understanding different aspects of COVID-19.

## Contribution

The study validates in silico predicted microRNAs in real-world clinical samples, highlighting their diagnostic potential despite conflicting predictions.

## Key findings

- Serum levels of miR-196a-5p, miR-98-5p, and miR-27a-3p were significantly elevated in COVID-19 patients compared to healthy controls.
- miR-196a-5p showed high sensitivity and specificity for diagnosing COVID-19.
- Inflammatory markers like IL-6 and HIF-1α were elevated in patients compared to controls.

## Abstract

COVID-19 continues to evolve. Thus, new therapeutic alternatives are needed. Some in silico analyses, including the role of microRNAs, have shown promising results, but they are based on computational simulations and predictions, so their consistency with real-world results may vary. This study investigated the potential of three miRNAs (miR-196a-5p, miR-98-5p, and miR-27a-3p) obtained from published bioinformatics results as promising biomarkers for different aspects of COVID-19 (susceptibility, severity, and therapeutic possibilities). The current study found that, serum levels of these miRNAs were significantly elevated in COVID-19 patients compared to healthy controls, particularly during the third wave patients which is contrast to that reported in In silico study regarding miR-98-5p. Additionally, serum levels of inflammatory markers such as interleukin − 6 (IL-6) and hypoxia inducible factor- 1 α (HIF-1α) were higher in patients compared to healthy control. miR-196a-5p exhibited high sensitivity and specificity for COVID-19 diagnosis. Finally, we highlight the necessity of validating microRNAs in patient serum, as our findings show that their differential expression is robust and independent of the initial predictive model. This study effectively translates a bioinformatics hypothesis into a verifiable biomarker through experimental validation regardless of the contradictory findings between hypothesis and the actual results.