MUTATION-CENTRIC GRAPH NETWORKS: INTEGRATING LOCAL AND DISTAL GENOMIC CONTEXT

Date

2026-4-13

Author

Hüseynov, Ramal

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

99
views

0
downloads

Cite This

Somatic mutations drive the transformation of normal cells into cancer, yet the vast majority occur within non-coding regions, where their functional consequences remain poorly under- stood. A key challenge is that non-coding mutations may regulate distant genes through the three-dimensional chromatin architecture of the genome, a dimension that linear sequence- based approaches entirely miss. We present MutationNetwork, a graph-based computational framework that models this regulatory landscape by integrating long-range chromatin interac- tion data with local genomic overlap and gene annotations. Using each somatic mutation as a seed, the framework iteratively traverses the chromatin interaction network to capture local and distal genomic contexts. By expanding the graph through overlapping and interacting intervals, the model identifies genes reachable at varying topological depths. A symmetric integer indexing scheme enables constant-time retrieval of interaction and overlap relation- ships, making the approach scalable to whole-genome cohorts. Applied to 560 breast cancer whole-genome sequences, we show that mutation features derived from the chromatin in- teraction network distinguish Triple-Negative Breast Cancer from Luminal A tumors more effectively than direct mutation-gene linear proximity alone. These results demonstrate that the three-dimensional regulatory context of somatic mutations carries biologically meaningful and computationally accessible information for cancer subtype stratification.

Subject Keywords

Long-range chromatin interactions, somatic mutations, driver genes, graphs, networks

URI

https://hdl.handle.net/11511/119123

Collections

Graduate School of Informatics, Thesis