Characterizing genetic effects on molecular phenotypes at the single-cell resolution across brain regions in the context of substance use disorder and HIV

About This Grant

Project Summary Millions of individuals are affected with substance use disorders (SUD), posing a significant burden on these individuals, their families, and communities. There is a substantial comorbidity between SUD and HIV infection. HIV is also a risk factor for SUD because of the increased use of opioid pain medications that may lead to opioid addiction. The Single-Cell Opioid Responses in the Context of HIV (SCORCH) consortium was formed to gain insights into cellular and molecular responses in different brain regions to SUD and HIV by collecting single-cell transcriptomic and epigenomic data in affected brain regions from hundreds of human donors, as well as from animal models. It has been observed that SUD and HIV comorbidity may exacerbate cellular dysfunction beyond the effects of each condition alone, and the data generated by the SCORCH consortium provide opportunities for a comprehensive characterization of cellular states across conditions including control, HIV, OUD, and HIV+OUD. Preliminary data show substantial heterogeneity in molecular phenotypes across samples with the same exposure, e.g., HIV and SUD. Our premise is that the identifications of genetic variants mediating the effects of exposures to HIV and SUD will offer a unique angle to understand how different cell types in different brain regions respond to the exposures, and such understanding through genetic heterogeneity among individuals can lead to novel insights and clinical applications. We will apply state-of-the-art integrative methods to investigate how genetic variants affect molecular phenotypes in different cell types across brain regions with different exposure. We will accomplish this goal through three specific aims. The first aim will analyze total read counts from a transcript/isoform or peak using Bayesian methods that explicitly model shared genetic effects to borrow information across cell types and brain regions to increase statistical power. We will perform eQTL, caQTL, and isoQTL analyses. We will also leverage the multi-omic data to infer gene regulation networks and conduct grQTL analysis. The second aim will consider allele-specific analysis to complement analyses based on total counts. We will then combine allele-specific results with total read count results. To further improve statistical power, we will integrate SCORCH data with external data sets, computationally predicted effect sizes for genetic variants, and transfer known QTLs. We will develop gene expression and chromatin accessibility imputation models to facilitate genome-wide association studies. We will work with the SCORCH team to share our results with the broader scientific community. This project will be co-led by Dr. Hongyu Zhao, Dr. Mark Gerstein, and Dr. Ke Xu, who have complementary expertise covering statistical genetics/genomics, computational biology, single-cell analysis, SUD genetics, and HIV research.