Hierarchical Cooperation Of Transcription Factors From Integration Research

The spatial structure of the genome is important to cellular functions,such as gene expression and regulation,cell differentiation and identity,and tissue development.Linear DNA,which is 2 m long in one human cell,is highly packed into chromatin and adapts to the size of the cell nucleus.For a given species,different cell genomes are the same,but gene expression is cellular and tissue specific.With the development of highthroughput sequencing technologies and new biological methods,such as highthroughput chromosome conformation capture(Hi-C),in situ Hi-C,have been developed and applied to study the spatial organization of chromatin in various human cells.The structure of genomic DNA can be compartmentalized into four hierarchical structures with different resolutions: chromosome territories(50–250 MB bases),A/B compartments(~5 MB bases),topologically associated domains(TADs)(~1 MB)or sub-TADs(0.1–1 MB),and chromatin loops(5–300 kb).Among these structures,the chromatin loop is the architectural basis of other higher structures.The chromatin loop can bring distal regulatory elements,such as enhancers in linear DNA,to the promoters of target genes in 3D space.At present,the recognition of enhancer-promoter interaction is mainly through Chromosome Conformation Capture technology,which has the disadvantages of high noise,low precision,long time and high costs.So,it is urgent and important to study the interaction pattern of enhancer and promoter based on many new computing methods.Although the chromatin loop has an important role in gene regulation and disease,the specific mechanism,especially the involved molecules,such as proteins,and the cooperation of transcription factors(TFs)participating in the chromatin loop,remains poorly understood.In the present study,we came up with the HidPET(Hierarchical and Dynamic Analysis of TF Cooperation with ChIA-PET and ChIP-Seq Data)method to study the hierarchy and dynamics of TF cooperation by integrating ChIP-Seq and ChIA-PET datasets.Took four human cell lines(K562?GM12878?HUVEC?MCF7)for example,based on ChIA-PET three dimensional data and chromatin states data to filter the enhancer–promoter interactions,which play a dominant role in chromatin interaction.Then scanned the enriched TFs by using a given TF PWM,took the protein-protein interactions as the 3D TF interaction to construct 3D interaction network.Next,based on TFs' binding sites in genome from ChIP-Seq to calculate the 1D TFs' correlation and then constructed 1D correlation network.After fused the 3D interaction network and 1D correlation network,the hierarchy and local network parameters are analyzed across four cell lines.Community and clique analysis revealed the hierarchical and dynamic features of synergistic cooperative TF interactions in regulating gene expressions and chromatin 3D architecture,we raised three specified TFs cooperation model and combined evolution,co-expression,methylation of genes to explain their function.