R-loopBase v2.0: species- and cell type-resolved quantitative R-loop maps and R-loop regulatory resources

Abstract

R-loops are non-canonical nucleic acid structures implicated in diverse cellular processes and diseases. We previously developed R-loopBase to address the pressing challenges of inaccurate R-loop maps and the limited understanding of the R-loop regulatory network. However, R-loopBase provided only human R-loop maps with categorical confidence and did not account for cell type specificity. These limitations, coupled with the emerging R-loop profiling technologies and advances in R-loop regulatory mechanisms, have motivated the development of R-loopBase v2.0 (https://rloopbase.nju.edu.cn). After uniformly re-analysing 358 R-loop mapping datasets generated by 16 technologies, we introduce a quantitative R-loop scoring framework that assesses genome-wide R-loop confidence in human, demonstrating strong concordance with known R-loop biology. We further extend the framework to specific cell types and mouse, fruit fly, and yeast. In parallel, we substantially expand the collection of R-loop regulators across species, enabling the inference of changes in R-loop homeostasis. Building on these expanded datasets, we implement several new interactive features and analytical tools, including an expanded multi-species genome browser, deep learning-based R-loop prediction, R-loop homeostasis analysis, and a streamlined data download portal. We anticipate that R-loopBase v2.0 will serve as an enhanced, up-to-date one-stop resource for the R-loop research community.