UM Data Science Research Seminar with MERLN

Schedule

LECTURE 1 (12:00 -12:20) - Q&A (12:20 - 12:30)

Speaker: Yağmur Doğay

Subject: Discovering new kidney frontiers with a proximal tubule map

Abstract
Chronic kidney disease (CKD) affects over 800 million people. A hallmark of CKD is the progressive decline in kidney function, which results in the accumulation of uremic toxins. These toxins contribute to the progression of CKD and its comorbidities, in particular cardiovascular diseases. Given that CKD patients take on average over 12 medications per day, there is an urgent need to better understand potential toxin-drug, and drug-drug interactions. This project aims to develop a high-quality, mechanistically detailed map of Proximal Tubule (PT) transport and metabolism in an automated way as possible. Physiological maps are detailed, computer and human-readable representations of complex biological processes. In this study, data collection is centered around manual curation and automated literature mining using INDRA. CellDesigner is used to construct the map. To ensure accessibility, interactive visualization, and community feedback, the finalized map is deployed on the MINERVA platform. The current PT map includes >35 mechanisms from >200 curated scientific articles. The map components have a detailed annotation and mechanistic visualization to clarify the functional details of the transport processes. This mechanistic kidney map provides a valuable framework to deepen our understanding of CKD progression by capturing complex biological processes, offering researchers a platform to test hypotheses, and ultimately accelerating and optimizing the development of novel therapies and interventions

LECTURE 2 (12:30 -12:50) - Q&A (12:50 - 13:00)

Speaker: Iga Skorupska

Subject: Computational modelling of cAMP signaling in the context of Alzheimer’s disease

Abstract
Cyclic adenosine monophosphate (cAMP) is a key signaling molecule controlling many neuronal processes, and its dysregulation is associated with neurodegenerative and neurological disorders. Understanding how cAMP dynamics are shaped by regulatory enzymes is essential for identifying therapeutic targets. In this study, we apply computational modeling and simulation-based analysis to investigate how distinct phosphodiesterase 4D (PDE4D) isoforms modulate cAMP signaling and downstream cAMP response element-binding protein (CREB) phosphorylation. CREB is a link between cAMP molecular signaling dynamics to gene-regulatory outcomes. Using an ODE-based mechanistic model, we simulate isoform-specific inhibition scenarios to predict their effects on cAMP dynamics and CREB activation profiles. Model predictions generate testable hypotheses for in vitro validation, enabling iterative model refinement through data-driven parameter updates. This integrative modeling framework bridges biological signaling data and computational inference, offering a quantitative approach to dissect PDE4D-mediated regulation. Future extensions could incorporate spatial compartmentalization and disease-specific assumptions, supporting predictive modeling of cAMP signaling in neuronal contexts and guiding data-informed PDE4D-targeted drug design.