New algorithm to reconstruct single-cell specific 3D genome architectures published in Nature Communications
The labs of Dr. Jianliang Qian, Jianrong Wang and Yu Zhang recently published their paper “Tensor-FLAMINGO unravels the complexity of single-cell spatial architectures of genomes at high-resolution” in Nature Communications.
The multi-scale 3D conformation of the human genome plays pivotal roles in coordinating different molecular processes in each cell. In addition, the 3D genome architectures are dynamic, with high degrees of cell-to-cell variability. Characterizing the heterogeneous single-cell specific 3D genome configurations is thus one of the fundamental problems in molecular biology and biomedical applications.
In this study, a new low-rank tensor completion framework is developed to decode the spatial genome structures in different single cells. The new algorithm, called Tensor-FLAMINGO, is designed with the advantage of overcoming the sparsity problem in the data. Thus, Tensor-FLAMINGO can accurately reconstruct single-cell specific spatial genome architectures at the highest resolution to date. This new mathematical strategy can help us to zoom into individual cells and to evaluate the spatial folding of DNA sequences in the 3D space. It enables quantitative analysis of cell-to-cell heterogeneity of our 3D genome configurations. Notably, the large-scale reconstructed chromosomal structures unravel the systems-level insights into the crosstalk among molecular processes, the emergent patterns of chromatin compartments, differential domain formations, single-cell specific long-range chromatin loops, and multi-way spatial hubs of gene regulation. The new Tensor-FLAMINGO algorithm may serve multiple scientific communities and have impacts on the understandings of the underlying principles of gene regulation and may help to pinpoint the functional genetic variants associated with complex human diseases.
The paper can be found here.
