ABSTRACT – PROJECT 1 Endometriosis is a chronic, estrogen-dependent, inflammatory disease that affects ~10% of reproductive age women, resulting in debilitating pelvic pain, infertility, and compromised quality of life. It is characterized by anatomically and phenotypically diverse lesions of endometrial-like tissue superficially on the pelvic peritoneum, the ovaries and deeply infiltrating into pelvic organs, with resulting neuroangiogenesis, fibrosis, adhesions, pelvic pain and infertility. The pathogenesis of endometriosis relies on complex interactions between endometrial, peritoneal mesothelial and connective tissue cells and activation of local immune cell responses. There is profound dysfunction of the innate and adaptive immune systems, associated with inefficient lesion clearance and pelvic and systemic inflammation. As clinical classifications of endometriosis are maladapted to the heterogeneity of disease expression, diagnostics as well as effective treatments are lacking. Thus, precise understanding of the cellular and molecular pathobiology of endometriosis is a critical prerequisite to improve disease classification and inform diagnostic and therapeutic interventions. The goal of Project 1 is to determine the contribution of the immune system to the pathobiology of endometriosis on a single cell level, and using a data-driven strategy to derive and molecularly characterize objective disease classification. In Aim 1, we will determine the cellular composition and functional attributes of endometriosis lesions, their surrounding peritoneal/serosal niches, and eutopic endometrium through the lens of transcriptomic signatures at single cell resolution. Our hypothesis is that lesions and their niche environments have unique and functionally relevant transcriptomic signatures. In Aim 2, we will determine the contribution of the local and peripheral immune system to the pathobiology of endometriosis leveraging CYTOF technology. We will test the hypothesis that the local and peripheral myeloid phagocyte systems are dysfunctional in women with endometriosis. Local and systemic immunological data will be integrated to identify immunological signatures of dysfunctionality and to differentiate endometriosis disease types, along with functional studies. Finally, in Aim 3, we will leverage unsupervised machine learning techniques to integrate single-cell assessment of endometriosis lesions, surrounding tissue, endometrium, the local and peripheral immune systems and clinical data into a cross-tissue predictive model of disease classification. Our integrated approach will leverage hundreds of existing, clinically well-annotated biospecimens in our well established Human Endometrial Tissue & DNA Bank and ongoing accrual through our extensive network of physician and surgeon collaborators. The impact of this study will be to derive a replete transcriptomic and proteomic taxonomy of endometriosis lesions, their niche environments, eutopic endometrium...