SCL: a lattice-based approach to infer 3D chromosome structures from single-cell Hi-C data

Area of Science:

• Genomics
• Computational Biology
• Structural Biology

Background:

• Single-cell Hi-C data present unique challenges, including zero-inflation and difficulties in quantifying DNA segment proximity.
• Existing computational tools for modeling 3D chromosome structures from single-cell Hi-C data are limited.

Purpose of the Study:

• To develop a novel computational method, single-cell lattice (SCL), for reconstructing 3D chromosome structures using single-cell Hi-C data.
• To address the specific characteristics of single-cell Hi-C data, such as zero-inflation.

Main Methods:

• Developed the single-cell lattice (SCL) computational method.
• Designed a specialized loss function and a 2D Gaussian function tailored for single-cell Hi-C data.
• Represented chromosomes as beads-on-a-string within a 3D cubic lattice, employing Metropolis-Hastings simulation and simulated annealing for structure optimization.

Main Results:

• SCL successfully reconstructs 3D chromosome structures that closely fit single-cell Hi-C data at 500 and 50 kb resolutions.
• Evaluated SCL against existing modeling software, demonstrating its efficacy.
• Identified patterns of trans-chromosomal contact beads and enriched topologically associating domains (TADs) for Lamin-B1 and H3K4me3.

Conclusions:

• SCL provides an effective computational approach for modeling 3D chromosome structures from single-cell Hi-C data.
• The method accurately captures genomic organization and facilitates the study of TADs and their associated proteins.