Landscapes for Cell State Transition Leveraging by Single-Cell Multi-Omics - The overall goal of this project is to develop novel mathematic methods and toolkits to connect cell fate transition and epigenetic regulation across tissues and diseases. Cell fate transition often occurs in organ development, tissue regeneration, and pathogenesis. Dysregulation of the cell fate transition can lead to abnormal development or diseases, such as type 2 diabetes, obesity, heart failure, and Alzheimer’s disease. Quantitively decoding how cell fate changes can provide novel mechanistic insight into organogenesis and tissue regeneration, and help identify new strategies for the treatment of human diseases. However, our knowledge of cell fate transition and its regulation is only the tip of the iceberg due to the impracticality of long- term tracing of cell transcriptomes. In the past decade, numerous single-cell atlases containing millions of cells in different tissues, organs, developmental stages, and biological conditions are routinely developed by consortia such as the Human Cell Atlas (HCA). These atlases provide an opportunity for the unbiased study of cellular dynamics and the regulation mechanism. The lack of computational methods presents a major knowledge gap in the understanding of cell dynamics and the regulation leveraging by those large reference atlases. To address this knowledge gap, we proposed a new concept of “reference-based cellular dynamic inference”, which is a novel strategy to automatically annotate the cell state transition in new datasets by learning from the appropriate reference, allowing us to easily perform comparative analysis among different tissues and disease conditions. In this project, we will pursue three parallel but complementary research directions: 1) to develop the first computational methods and toolkits for generating cell dynamics atlases and analyzing cell state transition based on the appropriate reference atlases; 2) to develop novel statistical models for studying epigenetic regulation of cell fate from single-cell multiomics data; 3) to generate the first dynamic reference landscapes of cell differentiation, such as cardiogenesis, hematopoiesis, and neurogenesis, and in- house landscapes of transdifferentiation. This project will be built on the foundation of our recent studies for the development of computational approaches to uncover cell state transition from single-cell transcriptomes in both homogeneous and heterogeneous cell populations and the studies for investigating the role of epigenetic regulation on cell fate transition. The proposed studies will generate advanced computational toolkits and broadly applicable dynamic reference atlases, which are expected to reveal profound mechanisms controlling cell state transition in health and disease. In the long term, the ability to build cell dynamics reference landscapes will open a new horizon to understand the diversity of cell fate through comparative analyses across tissues and diseases and enhance regenerative medicine.
