Alzheimer's disease (AD) is characterized by extracellular deposition of amyloid beta (Aβ), followed by intraneuronal accumulation of hyperphosphorilated tau, culminating with neurodegeneration. Understanding the responses of the cell populations affected by AD is crucial to develop novel therapeutic approaches. With this project, we aim to characterize the gene regulatory programs that define the disease-associated cellular states. To this end, we will leverage single-cell multiomic (RNA and ATAC) datasets from human AD patients (MIT ROSMAP Multiomics) and a relevant mouse model of amyloid pathology to identify disease-associated enhancers. The combination of RNA and ATAC will allow us to define gene regulatory networks in the diseased cell types and compare them across species. Using deep learning models, i.e. ChromBPNet, trained on the chromatin accessibility profiles of the identified genomic regions, we will decode the enhancer’s regulatory logic of both mouse and human disease-associated cell types. The output from these models will then be used as input in Ledidi pipelines for in silico design of synthetic elements to target the affected cells. All this work will be conducted under the supervision of Enric Llorens, affiliated to Karolinska Institute (Department of Cell and Molecular Biology).