Answers are grounded in this paper and cite their sources.
Noncanonical DNA secondary structures arise from unique sequence patterns, and their formation in vivo is modulated by local genomic context, including DNA supercoiling, chromatin state, and transcriptional activity. They contribute fundamentally to DNA replication, transcription dynamics, genome organization, gene regulation, genetic diversity, genome instability, disease pathogenesis, and aging. However, current computational tools for non-B DNA detection are limited in scope. Most predict only one or a few structural classes. None of the tools annotate the novel, structurally distinct and biologically significant motifs characterized in recent years, viz., A-philic DNA, extended-guanine Z-DNA (eGZ-DNA), adenine-cytosine motifs (AC-motifs), and G-triplex intermediates. Furthermore, no existing tool identifies genomic regions where multiple non-B DNA conformations overlap as hybrid domains or coclusters as structural hotspots. In this study we present the Non-B DNA Finder, a high-throughput computational framework that addresses these limitations. It is the first unified platform to integrate the detection of 11 structural classes and 24 subclasses of non-B DNA motifs, encompassing both the classical non-B DNA repertoire and novel motif types. For the first time, it enables genome-wide annotation of hybrid overlap regions and non-B DNA structural clusters. To this end, we applied the pipeline to 11 complete genomes spanning a broad GC range of 17.6–76.2%. A total of ∼15 million non-B DNA motifs were identified, corresponding to an average density of ∼6 motifs/kb (range: 0.76–15.5 motifs/kb). Overall motif density was positively correlated with GC content (r = 0.37). G-quadruplexes were the most prevalent structural class in GC-rich actinobacterial genomes, reaching 7604 motifs/Mb in Miltoncostaea marina. Contrastingly, curved DNA dominated AT-rich endosymbionts, accounting for >70% of all structures in Candidatus Carsonella ruddii and Buchnera aphidicola. Importantly, hybrid interaction analysis revealed that genomic regions rarely encode isolated non-B DNA conformations. In the human genome, G-Quadruplex → i-Motif (9.57%), G-Quadruplex → Slipped DNA (9.46%), and G-Quadruplex → R-loop (5.99%) were the most preponderant hybrid pairs. Structural cluster analysis identified loci harboring up to seven co-occurring structural classes within a 300 bp window. Non-B DNA Finder is freely accessible at https://nonbdnafinder.streamlit.app/, and the source code is available at https://github.com/VRYella/NonBDNAFinder/.
No takes yet. Share an insight, caveat, or question.
Yella et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: