Related Experiment Videos
Outcome and Exposure Polygenic Risk Scores Can Help Reduce Information Bias and Selection Bias in Regression
Maxwell Salvatore1,2, Ritoban Kundu2,3, Jiacong Du2,3
1Department of Epidemiology, University of Michigan, Ann Arbor, Michigan, USA.
Genetic Epidemiology
|June 23, 2026
Summary
Leveraging genetic data, specifically polygenic risk scores (PRS), can reduce bias in electronic health record (EHR) biobank data. PRS-informed imputation and sample weighting improve estimates of outcome-exposure associations.
Area of Science:
- Genetics
- Biostatistics
- Bioinformatics
Background:
- Electronic health records (EHRs) are rich data sources but suffer from missing data and selection biases.
- These biases arise from non-random patient visits and sampling methods, complicating accurate analysis.
- Genetic data, often comprehensively collected in biobanks, offers a potential solution to these data limitations.
Purpose of the Study:
- To investigate the utility of genetic data, particularly polygenic risk scores (PRS), in mitigating biases within EHR-linked biobanks.
- To compare the performance of PRS-informed imputation against traditional methods like PRS-uninformed imputation and complete case analysis.
- To evaluate these methods under various missing data mechanisms (MCAR, MAR, MNAR) and sampling schemes (random vs. biased).
Main Methods:
- Simulations were conducted to model different missing data patterns and sampling biases.
- Evaluated methods included PRS-informed imputation, PRS-uninformed imputation, and complete case analysis.
- Real-world data from the Michigan Genomics Initiative (MGI) was used to validate simulation findings against national benchmarks.
Main Results:
- PRS-informed imputation demonstrated a reduction in bias and root mean square error (RMSE), alongside improved coverage, especially under missing at random (MAR) conditions in random samples.
- In biased samples with MAR exposure-only missingness, weighted PRS-informed imputation yielded substantially lower percent bias (0.6%) and better coverage (89.1%) than weighted complete case analysis (9.4%; 74.3%).
- MGI data analysis indicated that PRS-informed methods produced estimates more consistent with national benchmarks compared to unweighted complete case analysis.
Conclusions:
- Genetic data, when combined with sample weighting, can effectively reduce bias in association estimates derived from biobank data.
- Researchers are encouraged to utilize PRS for imputation and survey weighting methods for sample weighting when estimating outcome-exposure associations in target populations.
- The benefits of these PRS-informed approaches may vary depending on the specific outcome and data structure.
Related Concept Videos
01:23Types of Biopharmaceutical Studies: Controlled and Non-Controlled Approaches
Biopharmaceutical studies constitute a vital field aiming to enhance drug delivery methods and refine therapeutic approaches, drawing upon diverse interdisciplinary knowledge. In research methodologies, the choice between controlled and non-controlled studies significantly influences the study's reliability and accuracy.
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast, controlled...
Non-controlled studies, commonly employed for initial exploration, lack a control group, rendering them susceptible to biases and external influences. In contrast, controlled...
01:11Genome-wide Association Studies-GWAS
Genome-wide association studies or GWAS are used to identify whether common SNPs are associated with certain diseases. Suppose specific SNPs are more frequently observed in individuals with a particular disease than those without the disease. In that case, those SNPs are said to be associated with the disease. Chi-square analysis is performed to check the probability of the allele likely to be associated with the disease.
GWAS does not require the identification of the target gene involved in...
GWAS does not require the identification of the target gene involved in...
01:29Pharmacogenomics: Identification of New Drug Targets
Advances in genomics have profoundly influenced drug discovery by increasing both the speed and accuracy of pharmaceutical development. Pharmacogenomics, which examines how genetic variation influences drug response, facilitates the identification of novel therapeutic targets and enables patient stratification for personalized treatment. These strategies contribute to improved drug efficacy, minimized adverse effects, and more efficient clinical trial design.Mapping genetic differences...
01:18Polygenic Traits
When more than one gene is responsible for a given phenotype, the trait is considered polygenic. Human height is a polygenic trait. Studies have uncovered hundreds of loci that influence height, and there are believed to be many more. Due to the high number of genes involved, as well as environmental and nutritional factors, height varies significantly within a given population. The distribution of height forms a bell-shaped curve, with relatively few individuals in the population at the...
01:18Polygenic Traits
When more than one gene is responsible for a given phenotype, the trait is considered polygenic. Human height is a polygenic trait. Studies have uncovered hundreds of loci that influence height, and there are believed to be many more. Due to the high number of genes involved, as well as environmental and nutritional factors, height varies significantly within a given population. The distribution of height forms a bell-shaped curve, with relatively few individuals in the population at the...
01:29Bias in Epidemiological Studies
Biases can arise at various stages of research, from study design and data collection to analysis and interpretation. Recognizing and addressing these biases is essential to ensure the validity and reliability of epidemiological findings.Broadly speaking, biases in epidemiology fall into three main categories: selection bias, information bias, and confounding. A more detailed description of possible biases is: