Plasma proteome mendelian randomization and network pharmacology reveal therapeutic targets for thyroid disorders.

[INTRODUCTION] Thyroid disorders, including hypothyroidism, hyperthyroidism, and thyroid cancer, impose a substantial global health burden. Existing treatments face limitations due to adverse effects and incomplete efficacy, highlighting the need for innovative therapeutic strategies informed by genetic and molecular insights.

[METHODS] We integrated Mendelian randomization (MR) with network pharmacology to systematically prioritize druggable targets. Genetic correlations between different thyroid disorders were evaluated using linkage disequilibrium score regression analysis. Plasma protein quantitative trait loci from 4907 plasma proteins were leveraged as instrumental variables in MR analyses across two independent cohorts. Bayesian colocalization validated shared causal variants. Network pharmacology methods encompassed constructing protein-protein interaction networks, conducting functional enrichment analyses, and identifying potential therapeutic compounds via the DSigDB database. Docking and dynamics simulations assessed binding and stability, while PheWAS assessed off-target effects.

[RESULTS] The LDSC analysis identified notable genetic correlations of hypothyroidism with hyperthyroidism (Rg = 0.167, P = 0.017), as well as hyperthyroidism with thyroid cancer (Rg = 0.286, P = 0.033). MR and colocalization identified seven causal proteins: IL2RB, CDH1, FGF19 (hypothyroidism); PSAPL1 (hyperthyroidism); DCP1B, SPRN, RPS6KA6 (thyroid cancer). Drug prediction prioritized compounds such as BI-2536 (binding energy: -9.5 kcal/mol with RPS6KA6) and deoxycholic acid. PheWAS confirmed minimal pleiotropic risks.

[CONCLUSIONS] By synergizing genetic epidemiology with network pharmacology, this study delineates shared genetic architecture among thyroid disorders and nominates seven high-confidence targets with therapeutic potential. The integrative framework advances precision medicine by bridging causal plasma protein identification, mechanistic pathway mapping, and drug repurposing, offering a blueprint for multi-omics-driven drug discovery in endocrine pathologies.