This paper discusses the use of summarized data from genome-wide association studies (GWAS) in Mendelian randomization (MR) to estimate the causal effect of a risk factor on an outcome. The authors present two methods for combining regression coefficients from multiple genetic variants: an inverse-variance weighted (IVW) approach and a likelihood-based method. They compare the bias and efficiency of these methods to those based on individual-level data through simulation studies, investigating the impact of gene-gene interactions, linkage disequilibrium, and weak instruments. The results show that both methods provide similar estimates and precision to individual-level data methods, even with gene-gene interactions, but overstate precision when variants are in linkage disequilibrium. The paper also includes an example using published data on five genetic variants associated with low-density lipoprotein cholesterol (LDL-C) to estimate the causal effect of LDL-C on coronary artery disease (CAD). The estimated reduction in CAD risk from a 30% decrease in LDL-C is 67% (95% CI: 54% to 76%). The authors conclude that MR investigations using summarized data from uncorrelated variants are efficient, but assumptions cannot be fully assessed without individual-level data.This paper discusses the use of summarized data from genome-wide association studies (GWAS) in Mendelian randomization (MR) to estimate the causal effect of a risk factor on an outcome. The authors present two methods for combining regression coefficients from multiple genetic variants: an inverse-variance weighted (IVW) approach and a likelihood-based method. They compare the bias and efficiency of these methods to those based on individual-level data through simulation studies, investigating the impact of gene-gene interactions, linkage disequilibrium, and weak instruments. The results show that both methods provide similar estimates and precision to individual-level data methods, even with gene-gene interactions, but overstate precision when variants are in linkage disequilibrium. The paper also includes an example using published data on five genetic variants associated with low-density lipoprotein cholesterol (LDL-C) to estimate the causal effect of LDL-C on coronary artery disease (CAD). The estimated reduction in CAD risk from a 30% decrease in LDL-C is 67% (95% CI: 54% to 76%). The authors conclude that MR investigations using summarized data from uncorrelated variants are efficient, but assumptions cannot be fully assessed without individual-level data.