Dataset Information

PASTRY: achieving balanced power for detecting risk and protective minor alleles in meta-analysis of association studies with overlapping subjects.

ABSTRACT:

Background

Meta-analysis is a statistical method that combines the results of multiple studies to increase statistical power. When multiple studies participating in a meta-analysis utilize the same public dataset as controls, the summary statistics from these studies become correlated. To solve this challenge, Lin and Sullivan proposed a method to provide an optimal test statistic adjusted for the correlation. This method quickly became the standard practice. However, we identified an unexpected power asymmetry phenomenon in this standard framework. This can lead to unbalanced power for detecting protective minor alleles and risk minor alleles.

Results

We found that the power asymmetry of the current framework is mainly due to the errors in approximating the correlation term. We then developed a meta-analysis method based on an accurate correlation estimator, called PASTRY (A method to avoid Power ASymmeTRY). PASTRY outperformed the standard method on both simulated and real datasets in terms of the power symmetry.

Conclusions

Our findings suggest that PASTRY can help to alleviate the power asymmetry problem. PASTRY is available at https://github.com/hanlab-SNU/PASTRY .

SUBMITTER: Kim EE

PROVIDER: S-EPMC10790263 | biostudies-literature | 2024 Jan

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

PASTRY: achieving balanced power for detecting risk and protective minor alleles in meta-analysis of association studies with overlapping subjects.

Kim Emma E EE Jang Chloe Soohyun CS Kim Hakin H Han Buhm B

BMC bioinformatics 20240112 1

<h4>Background</h4>Meta-analysis is a statistical method that combines the results of multiple studies to increase statistical power. When multiple studies participating in a meta-analysis utilize the same public dataset as controls, the summary statistics from these studies become correlated. To solve this challenge, Lin and Sullivan proposed a method to provide an optimal test statistic adjusted for the correlation. This method quickly became the standard practice. However, we identified an un ...[more]

PMID: 38216869

Similar Datasets

Project description:BACKGROUND:Genome-wide association studies (GWASs) have revealed relationships between over 57,000 genetic variants and diseases. However, unlike Mendelian diseases, complex diseases arise from the interplay of multiple genetic and environmental factors. Natural selection has led to a high tendency of risk alleles to be enriched in minor alleles in Mendelian diseases. Therefore, an allele that was previously advantageous or neutral may later become harmful, making it a risk allele. METHODS:Using data in the NHGRI-EBI Catalog and the VARIMED database, we investigated whether (1) GWASs more easily detect risk alleles and (2) facilitate evolutionary insights by comparing risk allele frequencies of different diseases. We conducted computer simulations of P-values for association tests when major and minor alleles were risk alleles. We compared the expected proportion of SNVs whose risk alleles were minor alleles with the observed proportion. RESULTS:Our statistical results revealed that risk alleles were enriched in minor alleles, especially for variants with low minor allele frequencies (MAFs < 0.1). Our computer simulations revealed that > 50% risk alleles were minor alleles because of the larger difference in the power of GWASs to differentiate between minor and major alleles, especially with low MAFs or when the number of controls exceeds the number of cases. However, the observed ratios between minor and major alleles in low MAFs (< 0.1) were much larger than the expected ratios of GWAS's power imbalance, especially for diseases whose average risk allele frequencies were low, such as myopia, sudden cardiac arrest, and systemic lupus erythematosus. CONCLUSIONS:Minor alleles are more likely to be risk alleles in the published GWASs on complex diseases. One reason is that minor alleles are more easily detected as risk alleles in GWASs. Even when correcting for the GWAS's power imbalance, minor alleles are more likely to be risk alleles, especially in some diseases whose average risk allele frequencies are low. These analyses serve as a starting point for future studies on quantifying the degree of negative natural selection in various complex diseases.

Dataset Information

PASTRY: achieving balanced power for detecting risk and protective minor alleles in meta-analysis of association studies with overlapping subjects.

Background

Results

Conclusions

Publications

PASTRY: achieving balanced power for detecting risk and protective minor alleles in meta-analysis of association studies with overlapping subjects.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets