# A robust and powerful GWAS method for family trios supporting within-family Mendelian randomization analysis

## TL;DR

This paper introduces a new GWAS method for family trios that improves accuracy and reduces bias in genetic studies.

## Contribution

The novel FT-SEM method uses structural equation modeling to incorporate parental and offspring phenotypes and genotypes.

## Key findings

- FT-SEM improves estimation accuracy and test power while controlling bias and type I error rate.
- Using family trios reveals that dynastic effects and population stratification distort GWAS results.
- Combining FT-SEM with summary data shows biases in BMI effects on nicotine and alcohol consumption.

## Abstract

Effect size estimates in genome-wide association studies (GWAS) and Mendelian randomization (MR) studies for independent individuals may be biased due to dynastic effect (DE) and residual population stratification (RPS). Existing GWAS methods for family trios effectively controlled such biases, while only using parental and offspring’s genotypes and offspring’s phenotype, and not incorporating parental phenotypes, which causes loss in estimation accuracy and test power. Therefore, we proposed a novel GWAS method based on structural equation modelling for family trios, denoted by FT-SEM. FT-SEM simultaneously uses parental and offspring’s genotypes and phenotypes. Simulation results demonstrate that FT-SEM substantially improves estimation accuracy and test power while controlling bias and type I error rate. Using family trios from Minnesota Center for Twin and Family Research (MCTFR), we found that DE and RPS greatly distort the results only based on independent individuals, and FT-SEM effectively corrects such biases. Combining the GWAS results from MCTFR with existing summary data, we performed several two-sample MR analyses. We observed that the effects of BMI on nicotine, alcohol consumption and behavior disorder were due to bias rather than causality. Our findings underscore the necessity of using families to validate the results of GWAS and MR, and highlight FT-SEM’s advantages.