PROJECT SUMMARY The capacity for muscle regeneration decreases markedly with aging. While regeneration is led by muscle stem cells (MuSC), complex age-related changes in the skeletal muscle extracellular matrix (ECM) provide potent signals that drive aberrant lineage specification. The complexity of the interactions between aging MuSC and their environmental niche defined by biomechanical, architectural, and dynamic changes in the ECM suggests a data-driven analysis can elucidate underlying mechanisms, increase our fundamental understanding of aging and stem cell biology, and point to novel therapeutic strategies. In this research, -omics data (i.e., single cell RNA-seq and imaging flow cytometry assessments of myogenic markers) obtained from cells cultured onto substrates of varying elasticity and cell-adhesion will be used to probe signaling pathways including mitochondrial/metabolic signaling pathways in cultured MuSCs. We propose that the implementation of machine learning/artificial intelligence (ML/AI) paradigms represents a critical next step for integrating multi- layer -omics datasets and building predictive models that will more comprehensively elucidate stem cell responses to the extrinsic biophysical environment. The overarching goal of this Supplement is to test the central hypothesis that Biological data and domain knowledge relating to muscle aging can be embedded in a framework of Bayesian optimization will allow for elucidating mechanisms and accurately predicting regenerative responses. This central hypothesis will be tested by conducting three specific aims: Specific Aim 1. To prepare -omics data for ML models: Curate datasets, identify and impute missing data, compile metadata, and pre-process data to quantify descriptors used in model building. Adopt data management protocols associated with best practices. Specific Aim 2. To perform benchmark ML modeling with Bayesian optimization: Identify environmental variables (ECM stiffness and composition, signaling molecules) and cellular characteristics (age, expression markers) that correlate with epigenetic signatures and myogenicity, then develop mechanistic ML models and estimate posterior distributions. Specific Aim 3. To broaden approaches to ML modeling and broaden researcher engagement in the biology of aging: CMU will host a hackathon with teams that combine students and researchers from regional universities and HBCU partners.