Subtyping complex phenotypes via constrastive learning by leveraging electronic health records

利用电子健康记录通过对比学习对复杂表型进行亚型分类

• 批准号: 10799083
• 负责人: Elior Rahmani
• 金额: $ 42.84万
• 依托单位: UNIVERSITY OF CALIFORNIA LOS ANGELES
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-09-22 至 2025-08-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10799083
• 关键词: Address; Asthma; Breast Cancer Patient; Characteristics; Classification; Clinical; Complex; Data; Databases; Dementia; Diabetes Mellitus; Diagnosis; Dimensions; Disease; Drug Interactions; Electronic Health Record; Equity; Estrogen Receptors; Ethnic Origin; Genetic; Genomics; Graves' Disease; Hashimoto Disease; Health Surveys; Healthcare; Heterogeneity; Hypertension; Hyperthyroidism; Hypothyroidism; Individual; Insulin; Learning; Location; Medical; Medicine; Mental disorders; Methods; Modeling; Obesity; Patients; Pattern; Phenotype; Population; Population Heterogeneity; Predisposition; Probability; Procedures; Process; Prognosis; Race; Reporting; Risk Factors; Source; Stratification; Structure; Surface; Techniques; Time; Training; Undifferentiated; Variant; autoimmune thyroid disease; cell type; clinical heterogeneity; clinical risk; clinically relevant; cohort; computer framework; disorder subtype; estrophilin; high dimensionality; improved; learning strategy; multidimensional data; neoplastic cell; novel; novel strategies; phenotypic data; portability; precision medicine; predictive modeling; public health relevance; racial population; response; success

Summary A critical step towards realizing the promise of precision medicine is the identification of biologically- and clinically-relevant disease subtypes. Disease subtypes are suspected yet unknown or not fully characterized for many conditions, including obesity, diabetes, hypertension, asthma, dementia, and psychiatric disorders. The existence of “phenotypic heterogeneity” has practical and clinical implications: undifferentiated cases of a disease may represent the action of a variety of underlying causal processes, each of which may have a different prognosis or respond to a different treatment. Existing phenotype subtyping methods predominantly rely on the idea that applying clustering or dimensionality reduction techniques to high-dimensional data from patients with a given condition may reveal explanatory patterns that correspond to disease subtypes. This implicitly assumes that biologically meaningful subtypes can be captured by the dominant axes of variation in the data. Yet, the most dominant sources of variation are expected to be independent of biologically meaningful subtypes in many settings. In this project, a novel contrastive learning method is proposed for learning a heterogeneity gradient of variation that is specific to cases of a given condition and cannot be found in matched controls. Electronic health records (EHR) and survey information from the rich All of Us database is expected to span the spectrum of clinical heterogeneity across common complex diseases, which can inform the proposed method about meaningful sub-phenotypic variation for many diseases. The subtypes identified will be evaluated within the All of Us database and replicated using three external EHR cohorts for subtype- specific genetic effects, clinical risk factors, and clinical trajectories. Finally, EHR-based models are notoriously known for their susceptibility to poor generalization on out-of-distribution data that represent locations, populations, medical practices, or other factors that were not represented in the training data. This challenge will be addressed by developing a domain generalization framework, which will allow learning disease subtypes that are generalizable across demographic characteristics, including location, ancestry, ethnicity, and race, which is essential to achieve equitable precision medicine and facilitate the integration of predictive models in healthcare pipelines.

总结 实现精准医疗承诺的关键一步是识别生物学上和生物学上的 临床相关疾病亚型。疑似疾病亚型尚不清楚或未完全表征 用于许多疾病，包括肥胖症、糖尿病、高血压、哮喘、痴呆和精神疾病。 “表型异质性”的存在具有实际和临床意义： 疾病可能代表各种潜在的因果过程的作用，其中每一个都可能有一个 不同的预后或对不同治疗的反应。现有的表型分型方法主要是 依赖于将聚类或降维技术应用于高维数据的想法， 具有给定状况的患者可以揭示对应于疾病亚型的解释性模式。这 隐含地假设生物学上有意义的亚型可以被变异的主导轴捕获， 数据。然而，最主要的变异来源预计是独立的生物学 有意义的子类型。在这个项目中，提出了一种新的对比学习方法， 学习特定于给定条件的情况并且不能被发现的变化的异质性梯度 在匹配的对照组中。电子健康记录（EHR）和来自丰富的All of Us数据库的调查信息， 预计将跨越常见复杂疾病的临床异质性谱，这可以为 所提出的方法对许多疾病的有意义的亚表型变异。已识别的亚型 将在All of Us数据库中进行评估，并使用三个外部EHR队列进行复制，用于亚型- 特定的遗传效应、临床风险因素和临床轨迹。最后，基于EHR的模型 众所周知，它们对表示位置的分布外数据的概括性较差， 人口、医疗实践或其他因素，这些因素在训练数据中没有表现出来。这一挑战 将通过开发一个领域泛化框架来解决，这将允许学习疾病 亚型，可概括的人口统计学特征，包括位置，祖先，种族， 种族，这是实现公平的精准医疗和促进整合预测 医疗保健管道中的模型。