Biological processes operate through molecular networks at the cellular level, and through cell?cell networks at the tissue/organ level. Deciphering the ?wiring? and ?rewiring? of these networks under healthy and pathological conditions is a fundamental yet challenging goal of biomedical research. The emergence of single-cell RNA sequencing (scRNA-seq) has presented an unprecedented opportunity to achieve this goal by enabling genome- wide quantification of mRNA in thousands of cells simultaneously and overcoming the heterogeneity problem of bulk omics data. However, deep analysis of scRNA-seq data is challenging because only a small fraction of the transcriptome of each cell can be captured. No sophisticated computational tools are available to systemically reverse engineer intracellular gene?gene (especially signaling) networks and intercellular cell?cell interaction networks from single-cell omics data. Signaling proteins and epigenetic factors are crucial drivers of network rewiring and are most likely druggable, making them ideal therapeutic targets. Unfortunately, it is often difficult to unbiasedly identify many of these drivers (hence known as hidden drivers) because they may not be genetically altered or differentially expressed at the mRNA or protein levels, but rather are altered by posttranslational or other modifications. We have developed systems biology algorithms to expose hidden drivers from bulk omics data for antitumor immunity, tumorigenesis, and drug resistance. However, it remains even more challenging to reveal cell type?specific hidden drivers from scRNA-seq data because of the ?dropout? effects. Using our established state-of-the-art scRNA-seq platform, we profiled >100,000 epithelial cells from mouse mammary gland. Our ultradeep scRNA-seq profiling identified new subsets of somatic mammary stem cells (MaSCs) and shed light on the long-standing debate over the identities of multipotent and unipotent MaSCs. Therefore, building upon our expertise in systems biology, our robust preliminary results, and our established collaborations with leaders in the fields of breast cancer and immunology, we propose to develop computational algorithms to reverse engineer intracellular gene-wise and intercellular cell-wise networks (Aim 1), determine cell type?specific hidden drivers and their network rewiring (Aim 2), from single-cell omics data, and translate findings toward biomarkers and therapeutics to improve patient care (Aim 3). We will use information theory and Bayesian modeling in the development of these algorithms. We will use MaSCs and our breast cancer models as a proof of concept. With the increasing affordability of single-cell omics technologies, our algorithms can have a significant impact on many fields of biomedical investigation. For example, delineation of network rewiring and of critical drivers in stem cells and their niches will provide vital insights into cancer metastasis and relapse, and lay the foundation for understanding and overcoming the resistance of tumors to immunotherapies. Network- inferred hidden drivers are potential nonmutant therapeutic targets, and network-based biomarkers have tremendous potential to better stratify patients for precision cancer medicine.
Deciphering the wiring and rewiring of molecular and cellular networks is a fundamental yet challenging goal of biomedical research. The goal of our project is to develop computational algorithms for the inference of intracellular and intercellular network rewiring and underpinning hidden drivers from single-cell omics data. Our algorithms will allow us to answer fundamental questions about stem cells and their niches in healthy and pathological conditions, and thereby lay the foundation for discovery and development of better prevention, diagnosis and treatment strategies for human diseases.