Reliable Genetic Correlation Estimation via Multiple Sample Splitting and Smoothing

In this paper, we aim to investigate the problem of estimating genetic correlation between two traits. Instead making assumptions about distribution effect sizes factors, propose use a high-dimensional linear model relate trait factors. To estimate correlation, develop generic strategy that combines sparse penalization methods and multiple sample splitting approaches. The final is determined by taking median calculations, resulting in smoothed reliable estimate. Through simulations, demonstrate our proposed approach accurate comparison naive plug-in methods. further illustrate advantages method, apply it real-world example bacterial GWAS dataset, specifically antibiotic resistant traits Streptococus pneumoniae. This application not only validates effectiveness method but also highlights its potential applications.