Characteristics Of Hi-C Derived Data For Chromatin3D Modeling

Nuclear function regulation has a close relationship with spatial organization of genomes. Recent developments in the chromosome conformation capture (3C) technique and its derivatives (e.g., Hi-C) enable us to probe genome-wide chromatin interactions, which can advance our understanding of the genomic3D structure and its regulation. However, integrating data from these disparate experiments is difficult because of their significant differences and discrepancies, such as depth of available data, resolution scale, credibility, and noise. Effectively reconstructing the spatial chromatin structure from chromatin interaction data has become challenging to computational biologists. Existing modeling methods depend on a specific dataset or technique with the disadvantages of poor commonality, poor practicability, and low resolution.In this study, we introduce a novel approach for the genome-wide chromatin structure prediction, which is capable of relaxing biases such as those from different sequencing depth in Hi-C experiments. This method is based on an identified parameter that represents the inherent characteristics of Hi-C derived data for chromatin regions. We implement an automated pipeline for chromatin structure prediction as a web service (http://ibi.hzau.edu.cn/3dmodel/) for users to model regions of interest. We validate our method by intra and inter cell-line comparisons among various chromatin regions in several human cell-lines. For intra cell-line comparisons, we compare our results with data from fluorescence in situ hybridization (FISH) experiments and find that active regions are more open than inactive regions. Additionally, the active regions show significantly higher spatial distance than inactive regions in the invested cell lines. For inter cell-line comparisons, our modeling results are highly consistent with previous work based on5C data. Finally, by incorporating other omics data (such as chromatin modifications and gene expression) into analysis, we investigate the potential role of spatial chromatin organization in genome functions to provide further insight into the biological function of the chromatin structure.