Globally, mental illness is responsible for the most years lived with disability, and is especially challenging to address due to the lack of measurable biomarkers from inaccessible brain tissue. Genetics offers an objective measure of natural biological variability among diverse populations, providing a cornerstone of precision medicine with especially great promise for psychiatry, both as a gene discovery tool for therapeutic targets and as a substitute biomarker. While the Psychiatric Genomics Consortium (PGC) has amassed and jointly analyzed large-scale case-control datasets, these and other genome-wide association studies (GWAS) are Eurocentric, and the generalizability of these studies to diverse populations is low with standard approaches. The candidate hypothesizes that differences in allele frequency and the correlation structure of genetic variants along the genome (i.e. linkage disequilibrium or LD) are the primary culprits of poor generalizability, and proposes to test and improve the accuracy of genetic risk prediction across diverse populations. The proposed study will: 1) quantify genetic risk prediction accuracy of schizophrenia and related disorders across diverse populations (N?100k cases, N?213k controls); 2) build novel statistical methods that model LD differences across populations to improve genetic risk prediction when GWAS results are available in one or more populations, and risk prediction is desired in a mismatched population; and 3) build a method tailored to recently admixed populations that jointly models the mosaic of ancestry structure and LD to improve genetic risk prediction accuracy. The proposed studies and training plan were carefully designed to confer expertise in three domains: 1) the genetics of psychiatric disorders, 2) statistical methods development, and 3) large-scale data analysis and tools. These skills are fundamental to the candidate?s goal of becoming a leading investigator using human genetics as a lens into the evolution of complex traits, particularly psychiatric disorders. In addition to research training, the candidate will take coursework, participate in regular seminars, attend workshops and conferences, and gain mentorship experience locally and in Africa. All research will be conducted in the Analytic and Translational Genetics Unit at MGH and the Broad Institute with mentorship from Dr. Mark J. Daly, an established and prolific leader in human genetics. Additional mentorship from leading experts, Drs. Ben Neale, Karestan Koenen, Eimear Kenny, Jordan Smoller, and Sekar Kathiresan, ensures exceptional guidance. Overall, the training environment is outstanding, the mentors and advisors are world-class, the proposed studies address a crucial and timely unmet need, and the additional skills developed during this award will undoubtedly provide a strong foundation for the candidate to establish independent leadership in population, statistical, and psychiatric genomics.
Psychiatric disorders are the cause of the most years lived with disability, are highly heritable, and do not have measurable biomarkers due to the inaccessibility of brain tissue. While genetic studies hold great promise for psychiatry, providing the opportunity to discover new therapeutic targets and as a substitute biomarker from aggregated genome-wide risk, the vast majority of these studies are Eurocentric, and it is unclear how well these findings generalize across diverse populations. We propose to develop novel statistical methods that improve the generalizability of genome-wide association studies across populations for biological insights into psychiatric disorders.
