BPS2026 – Development of a database to support the understanding and visualization of 4D genome states

Advances in Hi-C technology have greatly enhanced our understanding of three-dimensional (3D) genome structures in relation to cellular states. However, since Hi-C relies on chemical fixation, the extent to which it reflects chromatin dynamics in living cells remains unclear. To address this, we previously developed PHi-C (polymer dynamics simulation from Hi-C data), a polymer physics-based framework that reconstructs Hi-C contact maps with over 99% accuracy. PHi-C enables the transformation of two-dimensional Hi-C data into dynamic 3D structures and maps of physical properties across genomic regions. Building on this approach, we are currently developing a web-based database to facilitate the interpretation of “4D genome states,” which incorporate both spatial and temporal chromatin information. To achieve this, we have established a high-performance GPU computing platform that supports large-scale analyses of publicly available Hi-C data. This system allows comprehensive PHi-C analyses at 50-kb resolution across entire chromosomes. Furthermore, we are constructing a web interface that integrates dynamic 3D simulations with 4D genome property maps, enabling users to visualize and explore genome organization in both space and time. Looking ahead, we plan to design a more comprehensive user interface that incorporates integration with existing genome databases, thereby providing broader accessibility and richer insights into genome dynamics.