Characterization of DNA extrusion scenarios underlying aggregate Hi-C patterns by a new computational framework

Abstract

1Aggregate Hi-C maps centered on several genetic elements show significant DNA extrusion features. However, these features are missing from the component Hi-C images. To resolve this discrepancy, we present a new computational framework that reconstructs DNA extrusion scenarios (DESs) and explains the enhancement of DNA extrusion features by aggregation. We showed that the reconstructed DES is essential to understand the site interactions of the relevant genetic elements. For example, CCCTC-binding factor (CTCF) binding elements (CBE) with a specific motif direction are only affected by DNA extrusion activities initiated from one side, manifested as a directional bias. About half of promoters search for enhancers in a specific direction, as indicated by the corresponding DES. Moreover, since DNA extrusion does not pause at promoter or enhancer elements, direct promoter-enhancer contacts are not long-lasting; rather, promoters and enhancers are brought in proximity by continuous, dynamic DNA extrusion.