Understanding Diabetes Heterogeneity via Mining Multimodality Interconnected Data

通过挖掘多模态互联数据了解糖尿病异质性

• 批准号: 10644701
• 负责人: Ji (Carl) Yang
• 金额: $ 17.28万
• 依托单位: EMORY UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-05-01 至 2028-04-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10644701
• 关键词: Amputation; Behavioral; Biological Markers; Biotechnology; Blindness; Caring; Characteristics; Chronic Disease; Clinical; Clinical Data; Complex; Computer software; Computers; Data; Data Science; Data Set; Detection; Development; Diabetes Mellitus; Disease; Disease Progression; Early Diagnosis; Early Intervention; Economic Burden; Economics; Electronic Health Record; Electronics; Environmental Risk Factor; Epidemic; Genomics; Goals; Grain; Graph; Health; Healthcare; Healthcare Systems; Heart Diseases; Heterogeneity; Human; Individual; Information Networks; Institution; Kidney Diseases; Life Experience; Machine Learning; Measures; Methods; Mining; Modality; Modeling; Modernization; Molecular; National Institute of Diabetes and Digestive and Kidney Diseases; Neural Network Simulation; Non-Insulin-Dependent Diabetes Mellitus; Outcome; Pathologic; Patients; Pattern; Phenotype; Population; Prediabetes syndrome; Prevalence; Production; Records; Research; Research Personnel; Scientist; Social Interaction; Source; Structure; Surveys; Symptoms; System; Training; Translating; Translational Research; United States; United States National Institutes of Health; advanced analytics; analytical tool; clinical phenotype; clinical practice; clinically actionable; cohort; combinatorial; comorbidity; complex data; computer science; cost; cost effective treatment; data structure; design; diabetes management; diabetic; diabetic patient; digital; effective therapy; graph neural network; health data; high risk; improved; individual patient; individual variation; insight; machine learning model; multimodal data; multimodality; multiple omics; novel; open source; outcome prediction; patient response; personalized medicine; precision medicine; prototype; secondary analysis; sensor; social; social factors; therapy design; time interval; treatment effect; treatment strategy

Understanding Diabetes Heterogeneity via Mining Multimodality Interconnected Data Abstract. Diabetes is a prevalent and highly heterogeneous disease that incurs tremendous human, economic, and so- cial costs globally. Prediabetes and early-stage type 2 diabetes often do not have single strong indicates or symptoms, posing great challenges for early detection and intervention. Moreover, once into a later-stage, diabetic patients are at high risk of developing various health problems such as heart disease, vision loss and kidney disease, which further com- plicates effective healthcare and may eventually lead to consequences from blindness to amputations to limited social interactions due to mobility. Unfortunately, current subtyping of diabetes has failed to decouple such heterogeneity. Recent remarkable advances in biotechnology have led to a significant production of high throughput patient data such as electronic health records (EHRs), multi-omics, and structured surveys, providing tremendous promises to powerful quantitative approaches towards the understanding of diabetes heterogeneity. However, existing machine learning (ML) models ignore the higher-order interconnections among various disease variables, thus failing to differentiate complex fine-grained subtypes and extract subtle corresponding phenotypes– regarding specific combinations of disease vari- ables. Moreover, most existing studies focus on single sources of data such as clinical, molecular or behavioral, failing to discover integrative biomarkers towards even more effective disease detection, analysis and treatment. Often case, these methods also rely heavily on dataset-specific feature preprocessing and fail to transfer from one cohort to another. As a computer scientist aiming at bridging data science and diabetes management, I have developed a well-structured training pan in this proposed K25 project, and my primary goal is to develop a high-impact and practical ML system that can be used to perform precise detection, in-depth analysis and cost-effective treatment of diabetes. To fully decouple the heterogeneity of diabetes from complex patient data, I propose (1) a hyper-hetero-graph data structure (H2G) to facilitate the comprehensive representation of patients and deep identification of diabetic characteristics regarding the interconnections among various disease variables and (2) a specialized graph neural network model (H2GNN) capable of modeling H2G along with a temporal component to capture the full trajectories of disease progression and a self- clustering component to identify novel subtypes of diabetes. Leveraging the national All of Us dataset from NIH with EHRs, genomics and surveys of 329K+ patients (42K+ diabetic), I propose to (1) apply the model to clinical data (EHRs) towards precise early diabetes detection, (2) incorporate molecular data (genomics) towards in-depth diabetes patho- logical analysis, and (3) further incorporate behavioral data (surveys) towards personalized diabetes treatment design. Finally, we will iteratively evaluate our system and its discovered subtypes based on both All of Us and our independent local dataset NELL curated from Emory Healthcare System with multimodality data of 295K patients (39K diabetic). This project is consistent with NIDDK's commitment to translational research on chronic diseases, and aligns well with ADA's recent initiative towards diabetes precision medicine. Results of this project will also inform the development of powerful quantitative approaches for a broader spectrum of diseases.

基于多模态关联数据挖掘的糖尿病异质性研究 抽象。糖尿病是一种普遍存在的高度异质性疾病，其对人类、经济等造成巨大的危害。 全球成本。前驱糖尿病和早期2型糖尿病通常没有单一的强烈迹象或症状， 对早期发现和干预提出了巨大挑战。此外，一旦进入晚期，糖尿病患者处于 患上各种健康问题的风险很高，如心脏病、视力下降和肾脏疾病，这些疾病进一步包括 使有效的医疗保健复杂化，并可能最终导致失明、截肢、社会福利受限等后果。 由于流动性的影响。不幸的是，目前的糖尿病亚型未能消除这种异质性。 最近生物技术的显著进步导致了高通量患者数据的大量产生， 作为电子健康记录（EHR），多组学和结构化调查，为强大的 量化方法对糖尿病异质性的理解。现有的机器学习（ML） 模型忽略了各种疾病变量之间的高阶相互联系，因此无法区分复杂的 细粒度的亚型，并提取微妙的相应表型-关于疾病瓦里的特定组合， 能力。此外，大多数现有的研究集中在单一的数据来源，如临床，分子或行为，未能 发现综合生物标志物，以实现更有效的疾病检测、分析和治疗。通常情况下， 方法还严重依赖于特定于队列的特征预处理，并且无法从一个队列转移到另一个队列。 作为一名旨在连接数据科学和糖尿病管理的计算机科学家，我开发了一个结构良好的 在这个拟议的K25项目中，我的主要目标是开发一个高影响力和实用的ML系统， 可用于对糖尿病进行精确检测、深入分析和经济有效的治疗。完全解耦 从复杂的患者数据的糖尿病的异质性，我提出（1）超异质图数据结构（H2 G）， 促进患者的全面代表和糖尿病特征的深入识别， 各种疾病变量之间的相互联系和（2）一个专门的图形神经网络模型（H2 GNN）能够 的建模H2 G沿着与时间的组成部分，以捕捉疾病进展的完整轨迹和自我- 聚类组件，以识别糖尿病的新亚型。利用来自NIH的全国All of Us数据集， EHR，基因组学和329 K+患者（42 K+糖尿病患者）的调查，我建议（1）将模型应用于临床数据（EHR） 实现精确的早期糖尿病检测，（2）将分子数据（基因组学）纳入深入的糖尿病病理学， 逻辑分析，以及（3）进一步结合行为数据（调查）以实现个性化糖尿病治疗设计。 最后，我们将根据我们所有人和我们独立的子类型， 来自Emory Healthcare System的本地数据集NELL，具有295 K患者（39 K糖尿病患者）的多模态数据。 该项目符合NIDDK对慢性病转化研究的承诺，并与 ADA最近对糖尿病精准医疗的倡议。该项目的成果还将为发展 为更广泛的疾病提供强大的定量方法。