Three-dimensional genomic data interpretation and transcriptional regulatory element identification

DNA is highly compressed and packaged as chromatin within the nucleus of eukaryote cells. Processes like DNA replication, DNA repair and gene transcription are precisely regulated by chromatin conformation which contributes to the self-maintenance, differentiation, specificity and identity of different cell types. In this collaborative project, we will analyse chromatin conformation. Firstly, by developing new and more accurate tools and algorithms to analyse Hi-C data and novel ChiA-PET data to identify the three-dimensional transcriptional network and its regulation. Secondly, we will apply the new tools/algorithms in two mammalian cell lines including mouse C2C12 cells, a model to study muscle development and porcine IPECJ2 intestine cells a model for porcine feed efficient studies. With the rapid development of chromatin conformation capture technology, genomic interaction data increases rapidly. This challenges the modelling methods and comparison methods of Hi-C data because these studies are computational intense. In the first part of this project, we will use libraries generated by the eHi-C technology to develop an efficient and accurate tool for Hi-C analysis. This part has three objectives: 1) to design a normalization approach for multi-library lateral comparison; 2) to design an algorithm to identify chromatin loops; 3) to develop a tool to analyse data of a newly developed technique (this technique is temporarily called multi-Chip), which can annotate the functions of loops and identify multiple interactions rather than only bidirectional interactions which is a limitation of current tools. For the second part, with the help of the new tool, we will construct precision 3D transcriptional regulation networks in the mentioned cell lines and build models to explore the correlation between eQTL and hierarchical structure for animal research. It has recently been suggested to apply binless function for Hi-C signal processing. While binning is the current state of the art and also the fastest method, binless approaches will be helpful to find a more precise structure of the chromosomes. To obtain a more in-depth insight of chromatin conformation, its dynamic trend and the comparison with existing data and algorithms, both binless and binning procedures will be utilized in our tool.