Exploring the Genetic Architecture of Uterine Fibroids

Abstract

Uterine fibroids (UF) are common benign tumors of the uterus affecting majority of females over their lifetime and are the number one cause of hysterectomies in the United States. Genetics are known to influence UF risk from previous familial and genome wide association studies (GWAS). UF have a variety associated comorbidities and phenotypes whose directionality relationship with UF is unknown. To identify high genetic risk for UF individuals, I constructed a meta PRS using summary statistics from FinnGen and Biobank Japan and the program PRS-CSx and validated it using 10-fold cross validation (10x CV) in Electronic Medical Records and Genomics (eMERGE) Network and BioVU. A previous phenome wide association study (PheWAS) for fibroids identified hundreds of associated phenotypes in the clinical phenome. Using mendelian randomization (MR), I assigned directionality to the relationships discovered. UF were found to be causal towards several other genitourinary phenotypes such as endometriosis and ovarian cysts, neoplasms, such as benign neoplasms of ovary and skin cancer, and appears to share horizontal pleiotropy with breast cancer and other breast conditions. From RNA-seq, I identified seven differentially expressed gens (DEGs) had previously been implicated by GWAS studies. Many of the genes are oncogenic. From this project, we can conclude that PRS can be built and applied to UF for prediction modeling. We also observed transcriptional differences in UF that reflect cancerous cell conditions and find that UF is causal towards other neoplasms.