Deep Exploration of Enhanced Integrated Model Based on EM Algorithm in Biostatistical Analysis and Disease Association Research

This article delves into the application of EM algorithm to optimize integrated models in biostatistical analysis and disease association research. It introduces an enhanced integrated model and two improved methods, with a focus on identifying disease-related mutation sites through statistical models. The research reviewed the latest developments in background, importance, and genome-wide association analysis; Then, the basic knowledge of biology and linear mixed models were introduced, and the feasibility of their extended applications was verified through simulation experiments. The improved integration model proposed in this study combines the p-value and additional information of each mutation site, and uses the EM algorithm to estimate parameters and verify their rationality. In order to address the challenges of high-density additional information processing, a scalable integrated model has been proposed, which utilizes a large amount of additional information for parameter estimation to screen for important mutation sites. Simulation experiments show that the improved model performs excellently in terms of FDR, AUC, and statistical power, and has achieved good results in real-world data applications.