A1349
Title: Geometric metric for identifying long-range interactions in three-dimensional chromatin structure
Authors: Yunchae Jung - Chungnam National University (Korea, South) [presenting]
Insu Jang - Korean Bioinformation Center Korea Research Institute of Bioscience and Biotechnology (Korea, South)
Minsu Park - Chungnam National University (Korea, South)
Abstract: Three-dimensional genome reconstruction from Hi-C data provides a geometric perspective on chromatin structure that complements traditional contact intensity analysis. While interactions between genomically adjacent loci typically dominate contact matrices, identifying distal pairs that are spatially close in reconstructed space is essential for understanding large-scale chromatin organization. A geometric metric is formulated by integrating spatial distance in reconstructed space and linear genomic distance along the polymer. This metric captures the geometric properties of the data without relying on parametric decay functions and maintains stability across multiple resolutions. Simulation studies under diverse structural configurations demonstrate that the metric effectively highlights long-range interactions often masked by strong local contact patterns. Application to Hi-C datasets confirms the successful identification of structural features not detectable through contact intensities alone. By integrating reconstruction models with genomic distance information, the proposed method offers a robust tool for comparative analysis in contexts where long-range interactions play a critical role.
