Dissecting enhancer function through integrative genomics

PROJECT SUMMARY Enhancers are non-coding regulatory elements that directly involved in modulating cell-type-specific gene transcription. Dissecting enhancer function provides critical understanding of human phenotypes and diseases. Current studies investigate enhancers by profiling one or few of their enhancer features, such as transcription factor binding, histone modification, chromatin accessibility, non-coding mutation, enhancer transcription and target gene interaction. To reach a comprehensive evaluation of enhancer function, dynamics, and importance to human diseases, there is urgent need to jointly analyze enhancer features across human cell types. However, key barriers exist towards the goal: data heterogeneity, missing enhancer profiles and lack of functional evaluation consensus. Over the years, we have gained extensive experiences and skills in research areas including sequence data quality control, machine learning algorithm development as well as enhancer functional evaluation. The long-term goal of the laboratory is to develop and apply computational methods for analyzing and interpreting gene regulation in human diseases. Here, we propose to leverage our experiences and skillsets to develop computational methods in enhancer function dissection through integration of eight different genomic features from hundreds of human cell types. Specifically, we will focus on three main areas in the next five years to address the barriers. i) We will develop statistical methods to harmonize heterogeneous data across human cell types so that sequencing quantification of enhancer features will be comparable and reflect valid enhancer activities. ii) We will develop machine learning methods to fill the missing enhancer profiles for unmeasured cell types by learning information from known enhancer profiles, DNA sequences and target gene expression. iii) We will develop integrative strategies to build enhancer connectome (enhancer-enhancer and enhancer-promoter interactions) and prioritize cell-type-specific importance of enhancers based on connectome topologies. The findings from the proposed work will significantly advance our understanding of enhancer function and their impacts in human diseases.