文献速递组学、生物样本库与精准医学医学人工智能与信息学流行病学与生物统计

整合多组学与多变量纵向数据分析用于阿尔茨海默病动态风险评估

Integrative Multi-Omics and Multivariate Longitudinal Data Analysis for Dynamic Risk Estimation in Alzheimer's Disease.
PubMed 文献 · Stat Med · 2025-01-01

阿尔茨海默病(AD)的严重程度和进展评估需要整合多变量纵向神经心理测试和多组学数据。研究开发了一个联合建模框架,结合多组学因子分析(MOFA)进行降维和特征提取,以及多变量函数混合模型(MFMM)建模纵向结局,从而利用组学和纵向数据动态评估痴呆风险。通过广泛模拟研究和ADNI数据集的实际应用验证了模型的有效性。

要点

  • 多组学与纵向数据整合面临高维、异质性和复杂互相关等挑战。
  • 联合建模框架结合MOFA降维与MFMM纵向建模,实现AD动态风险评估。
  • 模型通过模拟研究和ADNI数据集实际应用验证有效性。

核心结论

  • 联合建模框架可整合多组学与纵向数据,用于AD痴呆风险的动态评估。

研究设计与样本

  • 研究开发了一种新的联合建模框架,结合多组学因子分析(MOFA)与多变量函数混合模型(MFMM),并通过模拟研究和ADNI数据集实际应用进行验证。

主要发现

  • 该整合联合建模方法能够利用组学和纵向数据进行痴呆风险的动态评估,并在模拟研究和ADNI数据集中验证了有效性。

局限性

  • 原文未报告局限性。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

Alzheimer's disease (AD) is a complex and progressive neurodegenerative disorder, characterized by diverse cognitive and functional impairments that manifest heterogeneously across individuals, domains, and time. The accurate assessment of AD's severity and progression requires integrating a variety of data modalities, including multivariate longitudinal neuropsychological tests and multi-omics datasets such as metabolomics and lipidomics. These data sources provide valuable insights into risk factors associated with dementia onset. However, effectively utilizing omics data in dynamic risk estimation for AD progression is challenging due to issues including high dimensionality, heterogeneity, and complex intercorrelations. To address these challenges, we develop a novel joint-modeling framework that effectively combines multi-omics factor analysis (MOFA) for dimension reduction and feature extraction with a multivariate functional mixed model (MFMM) for modeling longitudinal outcomes. This integrative joint modeling approach enables dynamic evaluation of dementia risk by leveraging both omics and longitudinal data. We validate the efficacy of our integrative model through extensive simulation studies and its practical application to the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset.

来源:Stat Med,查看原文。版权归原作者所有。

引用

  1. Guo Y, Zou H, Alam MS, et al. Integrative Multi-Omics and Multivariate Longitudinal Data Analysis for Dynamic Risk Estimation in Alzheimer's Disease. Stat Med. 2025. PMID: 40387018. doi:10.1002/sim.70105

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。