Abstract
This study aimed to systematically identify and annotate oropharyngeal cancer (OPC) genetic risk, construct OPC risk prediction models, and identify candidate drug targets for OPC treatment. The OPC genome-wide association study was conducted in the UK Biobank involving 208 OPC patients and 328,248 healthy individuals. Functional annotations were performed using the FUMA website and the Sclinker statistical framework. OPC risk prediction models were constructed using logistic regression. Mendelian randomization (MR) was employed to identify potential OPC-related drug target genes. We identified 43 OPC risk genetic variants with p < 1E-06, which mapped to 7 independent loci. We also revealed significant tissue-level OPC single nucleotide polymorphism heritability enrichments in heart, kidney, and muscle, and cell-type specific enrichments in B cells, T cells, endothelial cells, epithelial cells, and fibroblasts. We further constructed four OPC risk prediction models, and the fully-adjusted model demonstrated moderate discriminative ability in internal validation. Additionally, MR analysis identified 160 nominate OPC-associated candidate druggable genes using expression/protein quantitative trait loci as instrumental variables. Drugs targeting these genes, such as Daunorubicin, Gefitinib, and Ibrutinib had been approved as anti-cancer agents. These results advanced our understanding of the genetic architecture of OPC and provided insights into its risk prediction and potential drug targets, contributing to future development of improved OPC prevention and treatment strategies.