DECODING BRAIN DISORDERS ACROSS THE LIFESPAN: INTEGRATIVE GWAS FROM NEURODEVELOPMENT TO NEURODEGENERATION

Understanding the genetic architecture underlying complex traits and disorders requires integrative analyses across multi-omic and neuroimaging modalities. In this thesis, we employed a series of large-scale genome-wide association studies (GWAS) and post-GWAS analyses to dissect the genetic basis of neurodegeneration and neuropsychiatric conditions, with a focus on Alzheimer's disease (AD), cognitive decline, and Tourette syndrome (TS). In the second section, we began by leveraging neuropathology-confirmed AD data to perform sex-specific GWAS and uncovered novel risk loci, including BIN1 (female-specific), QRFPR, and SGCZ, implicating disrupted lipid metabolism, peripheral immune dysfunction, and hormonal regulation in AD pathophysiology. Post-GWAS integration with transcriptomics, phenome-wide association studies (PheWAS), and Mendelian randomization (MR) provided causal evidence linking AD genetic risk to lipid profiles and inflammatory biomarkers. In the third section, we explored brain structural alterations associated with cognitive decline, identified genetic variants influencing longitudinal MRI-based brain changes, and, using polygenic risk scores (PRS), MR, and single-cell transcriptomic enrichment, demonstrated that genetic liability for anemia and type 2 diabetes contributed to neurodegeneration. In the fourth section, we compared mega- and meta-analysis pipelines across international TS cohorts to identify brain structural differences between TS and controls. We integrated GWAS data from TS and the ENIGMA consortium to evaluate pleiotropy between TS and cortical/subcortical brain measures, revealing shared genetic influences and distinct neurobiological patterns across neuropsychiatric disorders. Together, these findings demonstrated the use of multimodal GWAS frameworks in disentangling the biological complexity of brain disorders.