Multiomics data integration methods to discover putative causal variants, genes and patient heterogeneity for Alzheimers disease

PROJECT SUMMARY Despite substantial progress in identifying susceptibility loci for Alzheimer's disease (AD), challenges remain in pinpointing causal variants, understanding molecular processes, and characterizing patient heterogeneity. Our parent award, R01AG076901, focuses on developing computational methods to analyze quantitative trait loci (xQTL) data to elucidate the genetic basis of AD. To date, we have developed novel methods, including summary statistics based quality control for fine-mapping, multi-context colocalization, transcriptome-wide association studies (TWAS), and launched the FunGen-xQTL project, which leverages cloud computing infrastructure (MMCloud) to analyze large-scale multi-omics data efficiently. This supplement seeks one year of cloud computing support to extend our work and apply our methods to new data, focusing on high-reward goals of systematically mapping context-specific xQTL to provide vast data resources for the research community, and enhancing cloud computing infrastructure in academic settings. We propose to apply novel methods developed in the parent project to dissect xQTL across various contexts such as age, sex, APOE status, and disease state. Utilizing MMCloud and AWS resources, we aim to perform large-scale computational analyses efficiently at a genome-wide scale for both cis- and trans-effects from the current stage of FunGen-xQTL, including multi- context integration of linear, interaction, and quantile analysis models for xQTL and GWAS integration. We will also incorporate deep learning and machine learning techniques for functional annotation and prediction of xQTL effects to uncover causal variants and molecular processes underlying AD. Additionally, we will develop user- friendly utility scripts and comprehensive documentation to facilitate broader adoption and collaboration within the research community. This initiative promises to generate valuable insights into AD etiology and pave the way for precision therapies.