Exploring Endocrine Disrupting Pathways Using Knowledge Graph and Network Biology

Endocrine disruptors, such as pesticides and additives contained in microplastics, can threaten both human health and the environment, particularly by disrupting endocrine systems and imitating or inhibiting hormonal signals, resulting in a wide range of harmful biological effects. To investigate their molecular mechanisms, a computational framework named GELATO (Gene Expression anaLysis with integrATed Omics) was developed, integrating transcriptomics data with network-based approaches to identify gene and protein interaction networks, impacted pathways, Adverse Outcome Pathways (AOPs), and biological processes influenced by these chemicals. The analytical workflow combines differential expression analysis with Protein-Protein Interaction (PPI) mapping, and machine learning techniques, including clustering and network community detection, to identify key molecular signatures and network modules. Knowledge graphs are employed to organize biological data and simplify pathway investigation. Furthermore, literature mining serves to enhance the analysis by integrating structured knowledge from existing research, providing biological context and supporting data interpretation. This integrative tool seeks to provide a comprehensive framework for investigating endocrine disruption at a molecular level while also promoting a systems-level knowledge of the biological effects and potential adverse outcomes.