Our goal is to identify potential novel drug targets of Alzheimer?s disease (AD) through gene networks comprising genes associated with AD and genes targeted by anti-dementia or experimental drugs. The total payment in 2017 for caring for individuals with AD or dementia is estimated at $279 billion in the U.S. Despite significant effort and investment in the past decade, drug development for AD has not been successful. Failures in clinical trials are often due to severe adverse effects or lack of efficacy that could be confounded by failures to properly stratify etiological AD subgroups. Recent genomic studies have advanced our knowledge of the genetics of AD and comorbid conditions like vascular and metabolic symptoms, which could illuminate pathogenesis of AD subgroups. The new knowledge provides unprecedented opportunities for genetic based strategies for drug target identification and patient heterogeneity stratification. In our recent work linking antipsychotic drugs with schizophrenia, we mapped drug-disease overlap by overlaying drug target genes with disease risk genes. This genetic approach enables better understanding of disease pathogenesis, which can shed light on actions of current drugs and reveal novel druggable pathways for unmet therapeutic needs. Here we propose to apply our methods to AD. Our first objective (Aim 1) is to use existing omic data and an established network medicine approach (based on protein-protein interactions, which focuses on gene networks rather than individual genes) to determine which disease risk genes are targeted by potential drugs for AD (e.g., AChE inhibitors, NMDA antagonists, statins, NSAIDs, and anti- diabetic, ?-amyloid and tau agents). We will also identify untargeted disease genes to inform new drug discovery or repurposing. Drugs that interact with disease networks will be identified, including polypharmacy drugs that target multiple disease pathways. Our second objective (Aim 2) is to integrate omic and big data to determine therapeutic or adverse effects of drugs and identify repurposing drugs using data resources like the FDA AERS and AMP-AD Knowledge Portal. These informatics analyses can help prioritize drug targets, thus increasing the chances of a novel drug?s success. Our third objective (Aim 3) is to support the translation of drug-disease gene networks into clinical significance using patient data in clinical studies. We hypothesize that multiple pathways of AD gene networks may correspond to clinical heterogeneity.
For Aim 3, we will construct gene networks of AD related phenotypes such as lipid, insulin, vascular or memory pathways to test whether polygenic risk scores of these different networks correlate with heterogeneity of clinical variables using patient genotype data and medical records. Identifying genetically stratified subgroups by individual genotypes in multiple pathways will be important for patient enrichment designs in clinical trials and is aligned with personalized medicine. This project uses a multipronged approach with goals to aid in understanding current drug mechanisms, to identify targets for drug discovery and repurposing, and to detect genetic heterogeneity of AD.
Alzheimer?s disease is a devastating neurodegenerative disorder, but progress in developing therapeutics has been slow. We propose to use state-of-the-art network medicine approaches and the latest knowledge of disease and drug target genetics to increase understanding of the genetic bases of the disease and drug- disease associations, thereby shedding light on the actions of current drugs and revealing novel drug targets for unmet therapeutic needs. As a proof of concept, we will develop protocols to facilitate the translation of genomics knowledge gained from our network analyses into clinical practice and will test the power of patient genotype data for predicting clinical outcomes including cognitive decline using data gathered from large clinical studies.
