Grant: Project Grant

Major Depressive Disorder (MDD) is a leading cause of suffering and morbidity worldwide. The discovery of effective treatments to MDD is hindered by our incomplete understanding of the molecular basis of the disease. Recent advances in genomic techniques allow for the interrogation of genome-wide gene activities and cis-regulatory elements (CREs) at the single-cell resolution. We recently generated 160K single-nucleus transcriptomes and 160K single-nucleus chromatin-accessibility epigenomes from the post-mortem brains of 35 MDD and 31 control subjects. These multi-omic data, collected from the same brains but not measured in the same nuclei, provide the source information for decoding the regulatory programs in MDD. To that end, we propose 3 aims: - Aim 1. Linking CRE to MDD genes by aligning the single-nucleus transcriptomes and chromatin accessibility data; - Aim 2. Spatial prediction of cells to identify susceptible genes for the spatial composition changes in MDD brains, compare the differences between male and female MDD brains, and detect abnormal cell-cell interactions in MDD; - Aim 3. AI genome language models to perform in-silico mutation experiments to assess the effects of real-world genetic variants implicated in the MDD population. We also propose in-vitro experimental validation to validate the top hits from each aim. The proposed project will significantly advance our understanding of how the CRE-mediated genes can potentially influence the connectivity between cortical layers of the brain in a cell-type-specific and sex-specific way. Through spatial single-cell modeling, we will construct a biological graph connecting DNA regulatory elements, genes, cell types, and spatial features to facilitate our in-silico gene treatment strategy. Our results will identify therapeutic entry points for MDD. The proposed method is broadly generalizable to integrating unpaired single-cell multi-omic data and of a great utility to the research community.