Estimation of offspring genetic risk scores using parental genotypes

Objective Our objective was to determine whether genetic risk scores (GRSs) of offspring can be accurately estimated from parental DNA. Methods Whole genome sequencing data from a cohort of forty-seven multi-generation Utah families were used to extract single nucleotide polymorphism (SNP) data at genetic loci associated with the following traits: birth weight (BW), fasting plasma glucose (FPG), blood pressure (BP), body mass index (BMI), height, and type 2 diabetes (T2D). Offspring GRSs for each trait were estimated from parental single nucleotide polymorphism (SNP) data and compared to actual offspring GRSs. We also assessed offspring GRS estimation using only one parent’s DNA to simulate scenarios when only one genetic parent is available. The primary outcome was the percent error of parental-derived estimated GRS for each trait. An a priori threshold of 10% error was chosen for estimated GRSs to be considered accurate. Results Forty-three families with an average of 8.9 ± 1.8 offspring (N = 454 offspring) had parental and offspring SNP data available for GRS calculations. Mean percent errors for estimated offspring GRSs were less than 10% for all traits except for FPG (10.5% ± 8.1%). Percent errors were not significantly different when offspring GRSs were estimated using only one parent’s DNA whether the missing parent was a father or mother. Mean percent error of GRSs decreased exponentially with increasing SNPs per trait, with diminishing improvement in percent error above 500 SNPs. Conclusion Parental genetic risk scores can be used to accurately estimate genetic risk scores of offspring. This proof of concept supports further exploration of parental genetic risk scores as a tool for prenatal fetal genetic risk stratification. Statements What’s already known about this topic? Genetic risk scoring is a tool to estimate the probability of development traits or conditions with complex, multifactorial inheritance. What does this study add? Offspring genetic risk scores can be accurately estimated using parental DNA. This proof of concept supports further exploration of parental genetic risk scores as a tool for prenatal fetal genetic risk stratification.