Simulating topological domains in human chromosomes with a fitting-free model.

Area of Science:

• Genomics
• Computational Biology
• Biophysics

Background:

• Understanding the three-dimensional (3D) organization of human chromosomes is crucial for gene regulation and cellular function.
• Existing models often rely on extensive experimental data, limiting predictive power for novel genomic regions or cell types.

Purpose of the Study:

• To develop a polymer model that simulates chromosome folding based on 1D genomic data and transcription factor binding.
• To assess the model's ability to predict the formation of topologically associating domains (TADs) and their boundaries.

Main Methods:

• Representing chromosomes as strings of beads, colored by 1D bioinformatic data (e.g., chromatin state, histone modifications).
• Simulating reversible binding of transcription factors to specific bead colors using molecular dynamics.
• Analyzing the spontaneous formation of loops, rosettes, and TADs in the polymer model.

Main Results:

• The polymer model successfully folded into complex structures, including TADs, driven by transcription factor interactions.
• Simulations accurately predicted the locations of most TAD boundaries when compared with experimental Hi-C data.
• The model operates without fitting to Hi-C data, making it a 'fitting-free' predictive tool.

Conclusions:

• The developed polymer model offers a powerful, predictive framework for understanding chromosome 3D organization.
• This fitting-free approach can forecast the 3D structure of genomic regions lacking experimental Hi-C data.
• Future refinements can incorporate additional transcription factors and binding sites for enhanced predictive accuracy.