Features of the pathogenetic mechanisms of endometriosis associated with endometrial receptivity

Endometriosis is a highly prevalent gynecological disease that is often associated with impaired fertility, one of its underlying mechanisms being altered endometrial receptivity. The aim of this study was to systematically assess the relationship between transcriptomic changes in endometriosis and the molecular processes determining the receptive state of the endometrium. For a hypothesis-free, system-level comparison, principal component analysis was applied, followed by evaluation of the predictive significance of the resulting components for endometrial receptivity status using machine learning methods and their biological interpretation through enrichment analysis. Differential expression analysis revealed limited similarity between the conditions at the level of individual genes, whereas the system-level machine learning approach identified three principal components as the most informative for predicting endometrial receptivity. Enrichment analysis demonstrated that these components are associated with biological processes critical for implantation, including extracellular matrix remodeling, focal adhesions, Wnt signaling, immune response, and ciliary function. Modeling results indicate that this approach not only confirms known mechanisms involved in the establishment of endometrial receptivity but also enables the identification of novel factors associated with both receptivity and endometriosis.