CICADA: clinical informatics and computational approaches for drug-repositioning of AD/ADRD

点击翻译按钮获取中文摘要

Project Summary This proposal seeks support for developing advanced clinical informatics and computational approaches for drug-repositioning for Alzheimer's disease (AD) and related dementias (ADRD). The proposed project directly addresses the areas of emphasis in PAR-20-156 to “develop computational methods such as artificial intelligence/machine learning to investigate new uses of FDA-approved drugs or candidate drugs from failed Phase II/Phase III clinical trials through analysis of multimodal data.” The overarching goals of this proposal are to develop novel clinical informatics and computational approaches for drug repositioning of AD/ADRD. Specifically, we will develop statistical methods and ontology technology to extract drug-repositioning signals from multidimensional data (e.g., pharmacy-linked genetic data and biobank data, historical trials, and EHR data). The proposed framework is novel because it integrates advanced statistical inference procedures with semantic technology for data-driven and reproducible drug repositioning for AD/ADRD. We have three aims: We have three specific aims: Aim 1: Develop signal detection methods using multi-modal data (pharmacy-linked genetic data, genetic and electronic health record (EHR) data, and BioBank data). Aim 2: Evaluate the efficacy and safety of candidate drugs via historical trials and EHR data. Aim 3: Develop novel semantic and natural language processing (NLP) methods for Knowledge Graph (KG) construction. The success of this project will lead to novel computational methods, KG, and software for facilitating drug repositioning for AD/ADRD based on multimodal data. If successful, the proposed method could identify novel drug repositioning signals and generate novel hypotheses for prevention and treatment intervention of treat AD/ADRD. Our project holds the promise of identifying novel drug repositioning signals. This project is novel for integrating evidence synthesis methods with signal detection methods using advanced multimodal modeling, and it is potentially transformative for advancing prevention and treatment for AD/ADRD.