Defense Date


Document Type


Degree Name

Doctor of Philosophy


Human Genetics

First Advisor

Kenneth Kendler

Second Advisor

Ayman Fanous

Third Advisor

Brien Riley

Fourth Advisor

Brion Maher

Fifth Advisor

Hermine Maes


Understanding the genetic foundations of schizophrenia and the resultant symptom manifestations is an important step as we work toward development of new prevention and treatment strategies. This work has sought better understanding of this disease through use of three subject cohorts and two studies using simulated data exploring features of complex disease. First, we probed the symptoms of schizophrenia in subjects of African and European ancestry drawn from the Genetic Association Information Network (GAIN) schizophrenia study and found significant differences between groups, particularly in affective symptoms. The genetic basis of symptom variation was then explored in a selection of candidate genes in two Irish samples, the Irish Study of High Density Schizophrenia Families (ISHDSF) and Irish Case-Control Study of Schizophrenia (ICCSS). We found a significant association of PAH with delusions, GABRB3 with hallucinations, and SNAP25 with both of these symptom factors. AKT1 alleles conferred greater Schneiderian symptoms, but dysbindin, MAOB, and SLC6A4 were not related to any symptom dimensions. Simulated data were used to probe the parameters necessary to detect susceptibility genes as modifiers in a scenario in which two disease groups with incompletely overlapping symptom profiles are examined together. The heterogeneous genetic underpinnings and variable symptom manifestation of schizophrenia make the findings from this study particularly relevant to this disease. Convergent lines of evidence implicating myelin and synaptic dysfunction in schizophrenia prompted us to test related gene networks for association with this disease in two populations, African-ancestry and European-ancestry, from the GAIN study. Some evidence supporting myelin-related genes in the etiology of schizophrenia was presented but only in the African-ancestry group. Epistatic (gene-gene) interactions may confer much greater disease risk than single-gene results would indicate, but their detection is often difficult. The final study included here explored two approaches to family-based epistasis detection under a range of epistatic models. The haplotype relative risk (HRR) approach yields greater power for detection under conditions of dominance, but the Cordell approach is more powerful under most other models. Together, these studies provide a modest advancement in our understanding of schizophrenia and the methodological avenues available for future studies of this disease.


© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

September 2009