Plasma proteomes and metabolism with genome-wide association data for causal effect identification in ovarian cancer

This study seeks to investigate the relationship between plasma metabolites or proteins and the risk of ovarian cancer through Mendelian randomization analysis and construct, while also developing a predictive model for resistance to chemotherapy.

Appropriate SNPs from GWAS data were selected as instrumental variables. Multiple methods, such as IVW, MR-Egger regression, and WME, were employed to investigate the causal relationship. A predictive model was established utilizing binary logistic regression based on the identified plasma protein genes. Four plasma metabolites and four plasma proteins were recognized as risk factors for ovarian cancer, whereas four plasma proteins were identified as protective factors. A predictive model for chemotherapy resistance was formulated with an AUC of 0.844 (p = 0.002).

Plasma metabolites and proteins may affect the risk of ovarian cancer and its resistance to chemotherapy. This study presents potential predictive factors and the underlying mechanisms influencing the onset, progression, and resistance of the disease.