Unraveling Genetic Loci in Addiction and Sex-differentiated Effects through Innovative Genomic Analysis

Restricted (Penn State Only)
- Author:
- Gao, Shuang
- Graduate Program:
- Biostatistics
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- August 15, 2024
- Committee Members:
- Arthur Berg, Professor in Charge/Director of Graduate Studies
Dajiang Liu, Chair & Dissertation Advisor
Dave Mauger, Major Field Member
Bibo Jiang, Outside Field Member
Laura Carrel, Outside Unit Member - Keywords:
- sex-differentiated analysis
statistical genetics
eQTLs
addiction
neurodegenerative disorders - Abstract:
- Recent advances in genetic studies have underscored the critical roles of non-coding variants in complex diseases like addiction and neurodegeneration. However, the specific functions of these variants often remain unclear due to challenges in detecting and interpreting their effects within the cellular context. To address this, our research includes two novel analytical frameworks, MAPBOX and JOBS Factor, to explore the genetic archtecture of substance use behaviors and neurodegenerative disorders through sex-differentiated and cell type-specific analyses. First, we propose MAPBOX (Meta-Analysis Plus Biobanks fOr seX-specific analysis), the trans-ethnic meta-analysis framework, leverages sex-combined studies and principal components to control for ancestry heterogeneity, enhancing the genetic discovery. This approach notably improves the detection of sex-differentiated genetic effects across diverse populations, overcoming the complexities and biases typically encountered in genetic association studies. We validated MAPBOX through extensive simulations and applied it to large datasets including the GSCAN and MVP, encompassing over 3.4 million individuals. Our findings demonstrate that MAPBOX not only efficiently manages type I error rates but also boosts the discovery of sex-differentiated genetic loci by 71%. Further conditional analysis and fine-mapping allowed us to identify biologically significant genes like PLCD3 and CRF1, with CRF1 showing potential sex-biased signaling linked to stress responses. Second, we introduce the JOBS Factor (JOint model viewing Bk-eQTLs as a weighted sum of Sc-eQTLs with cell type Factor) method integrates bulk and single-cell RNA sequencing data to refine expression quantitative trait loci (eQTL) analysis, addressing the sample size limitations and enhancing the detection of cell type-specific eQTLs. By combining bulk eQTL data from the Cortex region of the MetaBrain dataset with single-cell data from Bryois et al. and Fujita et al., we achieved substantial increases in eQTL SNP discovery and replication across datasets. Integrating JOBS Factor with GWAS data for addiction-related traits and neurodegenerative disorders resulted in significant improvements in colocalization with known GWAS loci and revealed key genes like CLU, PICALM, and RAB29, further elucidating the genetic frameworks of Alzheimer's and Parkinson's diseases. These integrative approaches not only pave the way for more in-depth analysis of sex-biased genetic architectures for complex traits but also narrow down the risk genes and enhance our understanding of the genetic basis underlying addiction traits in distinct brain cell types. This potentially offers novel insights for targeted therapeutic developments through considering both sex and cell-type specificity in genetic research.