Non-alcoholic fatty liver disease (NAFLD) is a chronic liver condition characterized by excessive fat accumulation in the liver. It is the most common liver disease among obese people and can progress from steatosis to non-alcoholic steatohepatitis (NASH), and more advanced stages, such as liver fibrosis, cirrhosis and hepatocarcinoma. Yet, the underlying cellular and molecular mechanisms that drive disease progression remain unclear. While the advances in single cell technologies have generated a large amount of single cell data for human liver, we still lack the developmental trajectories of all liver cell populations in a healthy tissue environment and during the progression of NAFLD. Here, by integrating a large amount of published single cell datasets using our newly developed computational methods, we will standardize cell type nomenclature from various studies and establish a development trajectory of liver single cells in a homeostatic state. This integrated reference map will be trained by a machine-learning model to enable auto-projection of single cells from NAFLD patients of different disease stages onto the cell developmental path. Moreover, we will further develop novel computational tools to investigate the role of alternative splicing in NAFLD, especially in myeloid cells. Our study will contribute to a deeper understanding of cellular heterogeneity, shedding light on the mechanisms that initiate and drive the progression of NAFLD.