# Systematic computational hunting for small RNAs derived from ncRNAs during dengue virus infection in endothelial HMEC-1 cells

## TL;DR

This study develops a computational method to detect small RNA fragments from non-coding RNAs during dengue virus infection in human endothelial cells.

## Contribution

A novel computational approach for detecting differentially expressed small RNA fragments using copy-number-corrected annotations.

## Key findings

- The method successfully identified differentially expressed miRNAs and sfd-RNAs in dengue-infected cells.
- hsa-mir-103a and hsa-mir-494 were linked to potential molecular targets CDK5 and PI3/AKT in dengue virus infection.
- The approach addresses key challenges like multi-mapping reads and ncRNA annotation.

## Abstract

In recent years, a population of small RNA fragments derived from non-coding RNAs (sfd-RNAs) has gained significant interest due to its functional and structural resemblance to miRNAs, adding another level of complexity to our comprehension of small-RNA-mediated gene regulation. Despite this, scientists need more tools to test the differential expression of sfd-RNAs since the current methods to detect miRNAs may not be directly applied to them. The primary reasons are the lack of accurate small RNA and ncRNA annotation, the multi-mapping read (MMR) placement, and the multicopy nature of ncRNAs in the human genome. To solve these issues, a methodology that allows the detection of differentially expressed sfd-RNAs, including canonical miRNAs, by using an integrated copy-number-corrected ncRNA annotation was implemented. This approach was coupled with sixteen different computational strategies composed of combinations of four aligners and four normalization methods to provide a rank-order of prediction for each differentially expressed sfd-RNA. By systematically addressing the three main problems, we could detect differentially expressed miRNAs and sfd-RNAs in dengue virus-infected human dermal microvascular endothelial cells. Although more biological evaluations are required, two molecular targets of the hsa-mir-103a and hsa-mir-494 (CDK5 and PI3/AKT) appear relevant for dengue virus (DENV) infections. Here, we performed a comprehensive annotation and differential expression analysis, which can be applied in other studies addressing the role of small fragment RNA populations derived from ncRNAs in virus infection.