G-quadruplexes (G4s) are non-canonical four-stranded structures and emerging as novel genetic regulatory elements. However, a comprehensive, multidimensional profile of the contribution of endogenous G4s instead of computationally predicted G4s (PQS) to human gene regulation is still lacking. Here, we analyzed all public G4 ChIP-seq data and identified ~400,000 human endogenous G4s, accounting for 0.4% of the genome, significantly lower than previously estimated (1%). Compared with PQS, we found that G4s were more structurally stable and enriched in proximal promoter regions. We demonstrated that G4s were largely tissue specific, evolutionarily conserved at sequence and structure level, especially for constitutively formed G4s, and depleted at RNA moleculars. Integrating sample-matched ATAC-seq and RNA-seq data, our results showed that constitutive G4s often marked highly expressed genes. To further uncover the underlying mechanism, we generated a quantitative epigenetic and transcription factors (TFs) binding profile of G4s, and observed that constitutive G4s mainly carried active regulatory epigenetic states and concerned with complex transcriptional regulation. Finally, we nominated G4s that may perform regulatory functions in disease and cancer context, providing the scientific community with a powerful starting point to explore their regulatory mechanisms and biological relevance.

性别；G-四链体