Beschreibung:
<jats:title>Abstract</jats:title>
<jats:p>R-loops are structures containing DNA-RNA hybrids and displaced single-stranded DNA, which are widely distributed across genomic regions. The generation and removal of R-loops is dynamically regulated by several factors including helicases and topoisomerases. Previously, we have identified high levels of R-loops associated with genomic instability in Embryonal Tumors with Multilayered Rosettes (ETMR) (Lambo et al. 2019), which is associated with a sensitivity to drugs targeting topoisomerases and DNA repair. However, it is unknown whether there are other pediatric cancers with high levels of R-loops, and whether R-loops in these tumors are also associated with genomic instability and sensitivity to specific treatments. Here, we have used real R-loops sequencing data sets, including ETMR DRIP-seq data, and the bioinformatic tool QmRLFS-finder (Wongsurawat et al. 2012), which can predict R-loops forming sequences (RLFSs) based on DNA sequence composition. By performing bioinformatic analyses of DRIP-seq data sets and integrating these with RLFSs from human reference genome simulations, we aim to analyze how reliable the different methods are to measure and predict R-loops and what role they have in cancer genomes. In ETMR, R-loops signals from DRIP-seq data were enriched around transcriptional start sites (TSS), comparable to R-loops distribution and enrichment in other published cell models. However, the overlap with RLFSs was limited, indicating that more validations with real DRIP-seq data from more ETMRs and other pediatric cancers are needed to validate the different methods. Taken together, until now, we developed methods to (semi-)quantify R-loops by combining real sequencing data and computational simulation approaches to investigate R-loops distribution in cancer genomes and to investigate whether R-loops are associated with breakpoints or other genetic aberrations in pediatric cancer entities.</jats:p>