Inferring Single-Cell 3D Chromosomal Structures Based on the Lennard-Jones Potential

Reconstructing three-dimensional (3D) chromosomal structures based on single-cell Hi-C data is a challenging scientific problem due to the extreme sparseness of data. In this research, we used Lennard-Jones potential reconstruct both 500 kb and high-resolution 50 A chromosome was represented by string or DNA beads put into 3D cubic lattice for simulations. 2D Gaussian function impute sparse contact matrices. We designed novel loss potential, in which ? value, i.e., well depth, indicate how stable binding every pair is. For bead pairs that have contacts their neighboring pairs, assigns them stronger stability. The Metropolis–Hastings algorithm try different locations beads, simulated annealing optimize function. proved correctness validness reconstructed evaluating models according multiple criteria comparing with 3D-FISH