Characterizing Chromatin Interactions

Characterizing chromatin interactions is a crucial step towards increasing our understanding of the three-dimensional chromatin. Chromatin spatial organization plays a critical role in genome functions, such as gene regulation, by allowing distal enhancers to influence the transcription of their target genes. Most existing methods to detect significant long-range chromatin interactions assume statistical independence between interactions, an assumption which is invalid at high resolution. In Chapter 2, we present HiC-ACT, an aggregated Cauchy test (ACT) based approach, to improve the detection of chromatin interactions by post-processing results from methods assuming independence. HiC-ACT demonstrates advantages in improving sensitivity with controlled type-I error. By leveraging information from neighboring chromatin interactions, HiC-ACT enhances the power to detect interactions with lower signal to noise ratio and similar (if not stronger) epigenetic signatures that suggest regulatory roles. We further show that the most significant HiC-ACT peaks overlap with more known enhancers than those from the state-of-the-art Fit-Hi-C/FitHiC2 method. Existing studies of chromatin conformation have primarily focused on potential enhancers interacting with gene promoters. By contrast, the interactivity of promoters per se, while equally critical to understanding transcriptional control, has been largely unexplored. In Chapter 3, we leverage promoter capture Hi-C data to identify and characterize cell type-specific super-interactive promoters (SIPs). Notably, promoter-interacting regions (PIRs) of SIPs are more likely to overlap with cell type-specific ATAC-seq peaks and GWAS variants for relevant blood cell traits than PIRs of non-SIPs. Further, SIP genes show enriched heritability of relevant blood cell trait(s). Importantly, this analysis demonstrates the potential of using promoter-centric analyses of chromatin spatial organization data to identify biologically important genes and their regulatory regions. As over 90% of GWAS variants lie in non-coding DNA, it is challenging to identify the causal variants, although critical to understanding genetic mechanisms underlying diseases. Chapter 4 proposes MARVIN (Multi-disease/trait Annotation of Regulatory Variants using Integrated Networks), a method to predict the impact of non-coding regulatory variants on effector genes relevant to groups of diseases/traits at varying resolution. By constructing disease-specific gene regulatory networks and simultaneously modeling related diseases, MARVIN has increased power to accurately predict causal variants.Doctor of Philosoph