Role of extracellular matrix in age-related declines of muscle regeneration

项目总结 随着年龄的增长，肌肉再生能力显著下降。而再生是由肌肉主导的 干细胞(MUSC)，骨骼肌细胞外基质(ECM)中与年龄相关的复杂变化，提供 驱动异常血统规范的强有力的信号。老龄化之间相互作用的复杂性 MUSC及其由生物力学、建筑和动态变化定义的环境生态位 ECM认为，数据驱动的分析可以阐明潜在的机制，增加我们的基本面 了解衰老和干细胞生物学，并指出新的治疗策略。在这项研究中，-组学 数据(即单细胞rna-seq和成象流式细胞术对肌源性标记物的评估)来自 培养在不同弹性和细胞黏附的底物上的细胞将被用来探测信号通路。 包括培养的MuSCs中的线粒体/代谢信号通路。我们建议实施 机器学习/人工智能(ML/AI)范例代表着将多个 层组学数据集和构建预测模型，以更全面地阐明干细胞 对外在生物物理环境的反应。 本附录的首要目标是检验生物数据和领域的中心假设 与肌肉老化有关的知识可以嵌入到贝叶斯优化框架中，从而允许 阐明再生机制并准确预测再生反应。这一中心假设将是 通过执行三个特定目标进行测试：特定目标1.为ML模型准备组学数据：策划人 数据集、识别和归类丢失的数据、编译元数据以及对数据进行预处理以量化描述符 用于模型构建。采用与最佳实践相关的数据管理协议。具体目标2.至 使用贝叶斯优化执行基准ML建模：确定环境变量(ECM刚度 和组成、信号分子)和细胞特征(年龄、表达标记物) 表观遗传特征和生肌性，然后建立机械ML模型并估计后验 分配。具体目标3.拓宽最大似然建模的方法并扩大研究人员在 衰老的生物学：芝加哥大学将举办一场黑客马拉松，由来自 地区大学和HBCU合作伙伴。

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.