CoMeT

Gene regulation drives metabolic responses to environmental cues, such as fasting. Yet, the molecular mechanisms of this metabolic rewiring in highly sensitive organs such as liver is still poorly understood. Nuclear metabolism impacts gene expression through allosteric interactions, provision of energy, building blocks and co-factor supply crucial for epigenetic regulation. Thus, nuclear metabolism stands as a critical yet unknown link to transcriptional regulation. We also largely do not know how nutrient availability is sensed and signal transduction interplays with nuclear metabolism in transcription. These knowledge gaps are partially due to limitations of the existing computational tools for integrating diverse omics data and ignore of nuclear metabolome. Notably, current computational tools struggle with specificity, relying on pre-existing knowledge, do not encompass metabolic compartmentalization and falter in predicting dynamics of transcriptional responses under complex physiological conditions. The interdisciplinary CoMeT project will develop novel Computational methods with the aim to unravel new mechanistic connections between nuclear Metabolism and Transcriptional regulation in adaptation to fasting. CoMeT synergies complementary expertise from three partners in computational data analysis (Lutter lab), machine learning and data integration (Büttner lab), nutrient signaling, and transcriptional regulation (Panasyuk lab). By systematically exploring nuclear metabolomics, understanding predictability based on nuclear vs total metabolome, and adapting computational approaches, CoMeT commits to illuminating connections between metabolism and transcriptional regulation. Through interdisciplinary collaboration, innovative methodologies, and comprehensive multi-omics datasets, CoMeT aspires to deepen our understanding of metabolic adaptation to fasting, with three major objectives. First, CoMeT will develop prediction tools for metabolic impact on gene regulations in chronic fasting (calorie restriction) by integrating in prediction models nuclear metabolome, transcriptome and epigenetic profiles of H3K4me3 at the nexus of nuclear metabolism and active transcription. Second, largely supported by preliminary work, we will conduct mechanistic studies in reproducible cell models employing innovative molecular tools to establish the role of nuclear metabolism and nutrient sensing PI3K-3 signaling in transcriptional rewiring in fasting. Finally, we will employ developed computational models and new datasets to make predictions of sex-dimorphic transcriptional regulators in responses to fasting in liver. Altogether, CoMeT will advance our understanding of gene regulation. It will also provide cutting-edge machine learning methods and computational tools that will be instrumental for better comprehension of metabolic orchestration in health and disease.