Modeling Chromatin Regulation of Cell Identity

• Applicant: Professor Dr. Fabian Müller
• Subject Area: Bioinformatics and Theoretical Biology ; General Genetics and Functional Genome Biology
• Term: from 2018 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 401841729

Although virtually all cells in an organism share the same genome, regulatory mechanisms give rise to hundreds of different, highly specialized cell types. These mechanisms are governed by chromatin signatures which determine DNA packaging, spatial organization, interactions with regulatory enzymes as well as RNA expression and which ultimately reflect the state of each individual cell. Various epigenetic, chromatin-associated marks can be charted using genome-wide DNA sequencing. In recent years, national and international efforts have created extensive maps describing epigenetic variability associated with different cell states. The increasing amount of available single-cell data allows for characterizing cell-to-cell heterogeneity and identifying relevant subpopulations of cells. Yet, while bioinformatics tools and standards for processing and interpreting data generated for individual epigenetic marks are widely established, computational approaches that simultaneously capture the genome-wide dynamics of multiple levels of regulation across cell states are just emerging. This proposal describes a roadmap towards a more integrated understanding of chromatin regulation: First, I will establish user-friendly analysis pipelines for the integrative processing of (epi-)genomic data. Multi-level features will be derived from DNA accessibility, chromatin conformation and RNA expression profiles across cell states in multiple systems associated with differentiation and disease, including hematopoiesis, adipogenesis, embryonic development and leukemia. Combining these features with current developments in machine learning, I will construct predictive models for inferring the regulatory states of cell populations and individual cells. Emphasizing model interpretability, I will derive feature signatures of cell identity. In a complementary approach, I will model the statistical dependencies of chromatin-associated features across cell types and (disease) conditions using machine learning and probabilistic graphical models. Finally, I will translate the statistical model implications into testable molecular hypotheses which will be validated in vitro and in vivo using perturbation experiments (genome and regulation editing) as well as technologies for determining physical and structural features of DNA. The described computational approaches provide a framework for an epigenetic definition of cell state through the systematic exploration of molecular patterns and dependencies. They are generally applicable across biological systems. The proposed project thus has the potential to significantly advance our understanding of the regulatory dynamics in stem cell differentiation as well as in disease onset and progression.

• DFG Programme: Research Fellowships
• International Connection: USA
• Host: Professor William James Greenleaf, Ph.D.