Date: September 14, 2025
Time: 1:00 pm to 3:00 pm
Track: Plenary - Presidential Symposium
Nearly 23 years after the arrival of the first complete human genome draft, our understanding of genome complexity is at once vast and only beginning. This knowledge has led to identification of many single gene associations with a variety of diseases. However, most neurological conditions seen in clinic have complex genetic underpinnings influenced by environmental factors or life experiences that may tip them toward health or disease. The confluence of genome data, computational capabilities, and methods for biological validation is making it possible to approach a real-world understanding of how genomic variation can determine individual vulnerabilities to diseases like ALS, late onset Alzheimer’s and Autism Spectrum. Moreover, these complex neurological diseases are often not limited to brain but involve multiple organ systems.
Large population genome wide association studies (GWAS) have yielded collections of risk variants for specific disorders that can now be examined in concert to generate polygenic risk scores (PRS). The challenge is now to use these disease-associated variants, which required tens of thousands of individuals to find, for PRS estimates relevant to individual patient care. When coupled with AI modeling, PRS can be used to learn more about gene interactions that influence phenotypic expression of disease and reveal new therapeutic targets. This session will explore the present and near future status of applications for PRS in precision medicine.
At the conclusion of this symposium, attendees should be able to:
This presentation will provide the background, current state, and future of using polygenic risk scores (PRS) to estimate disease risk and tailor preventive care. The presentation will focus on the lessons and advancements made within the electronic Medical Records and Genomics (eMERGE) Network to create and return PRS for 10 common diseases.
This presentation explores how polygenic risk scores (PRS) harness genetic insights from key mutations (SOD1, C9orf72, TARDBP, FUS) to pinpoint novel therapeutic targets for ALS, which is a devastating neurodegenerative disease with limited treatments. By integrating gene-targeted strategies and drug repurposing, we address challenges in clinical translation and chart new directions for effective ALS therapies.
This presentation will discuss how to map the complex landscape of neurological disorders through a combination of sequence-based AI frameworks with mutli-omics integrative modeling. Our approach dissects non-coding and coding genomic contributions to neurological disease and identifies dysregulated regulatory circuits to discover disease mechanisms and predict individual risk.
This presentation will discuss pathogenic processes leading to late onset Alzheimer’s Disease (LOAD) beginning as early as 20 years prior to the onset of mild cognitive impairment (MCI). Therefore, early implementation of therapeutic measures is crucial to the success of strategies to prevent or substantially delay progression of cognitive decline. Large scale GWAS have identified several hundred thousands of genetic risk variants for LOAD and have been used to generate polygenic risk scores (PRS) that predict genetic predisposition to the disease. The challenge is now to apply these variants to reliably assess individual risk of developing MCI or progressing from MCI to AD. We have taken a deep learning approach, Deep-PRS, that combines genetic, phenotypic and biological pathway interactions to stratify individual AD risk over time, as clinical indicators evolve. This meta-PRS method refines PRS by combining it with phenotypic traits and pathway-specific clustering that allows tailoring of clinical risk assessment to the patient’s unique genetic makeup. The potential for using such an integrated, machine learning approach longitudinally, to identify key risk factors and modulation of individual risk as interventions are implemented, will be discussed.