Efficiently controlling for case-control imbalance and sample relatedness in large-scale genetic association studies

By W. Zhou, J. Nielsen, L. Fritsche, Rounak Dey, M. Gabrielsen, B. Wolford, J. Lefaive, Peter Vandehaar, Sarah A Gagliano, A. Gifford and 9 more. Nature Genetics, 2017

1,245 citations recorded by the source corpus.

In genome-wide association studies (GWAS) for thousands of phenotypes in large biobanks, most binary traits have substantially fewer cases than controls. Both of the widely used approaches, the linear mixed model and the recently proposed logistic mixed model, perform poorly; they produce large type I error rates when used to analyze unbalanced case-control phenotypes. Here we propose a scalable and accurate generalized mixed model association test that uses the saddlepoint approximation to calibrate the distribution of score test statistics. This method, SAIGE (Scalable and Accurate Implementation of GEneralized mixed model), provides accurate P values even when case-control ratios are extremely unbalanced. SAIGE uses state-of-art optimization strategies to reduce computational costs; hence, it is applicable to GWAS for thousands of phenotypes by large biobanks. Through the analysis of UK Biobank data of 408,961 samples from white British participants with European ancestry for > 1,400 binary phenotypes, we show that SAIGE can efficiently analyze large sample data, controlling for unbalanced case-control ratios and sample relatedness. SAIGE (Scalable and Accurate Implementation of GEneralized mixed model) is a generalized mixed model association test that can efficiently analyze large data sets while controlling for unbalanced case-control ratios and sample relatedness, as shown by applying SAIGE to the UK Biobank data for > 1,400 binary phenotypes.

Subfields: Genetics, Rheumatology, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/4144262404079fc4044aad441e4efa2ea6a95c04

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Genetics

In Rheumatology

In Statistics and Probability

Citations in the corpus

A generalized linear mixed model association tool for biobank-scale data (2021, methodology): “SAIGE 14 is one of very few exceptions and is currently the most commonly used GLMM-based tool for biobank-scale data, because of its computational efficiency and well-calibrated test statistics of both common and rare variants for unbalanced binary traits.”
The power of genetic diversity in genome-wide association studies of lipids (2021, methodology): “Association analysis of the residuals for the majority of cohorts was carried out using a linear mixed-model approach in rvtests or with other similar software, including BOLT-LMM 45 , SAIGE 46 or deCode association software.”
Genome-wide enhancer maps link risk variants to disease genes (2021, methodology): “Quantitative traits were inverse rank transformed and associations were estimated using BOLT-LMM 53 for quantitative traits and SAIGE 54 for binary traits.”
Leveraging polygenic enrichments of gene features to predict genes underlying complex traits and diseases (2020, methodology): “As inputs to SuSiE, summary statistics for each region were obtained using BOLT-LMM 50 for quantitative traits and SAIGE 51 for binary traits, in sample dosage LD was computed using LDStore 61 and phenotypic variance was computed empirically.”
The Polygenic and Monogenic Basis of Blood Traits and Diseases (2020, methodology): “Thus, we performed a phenome-wide association study (PheWAS) for 456/574 rare variants using summary statistics for 529 well-represented clinical phenotypes from the UK Biobank cohort (https://www.leelabsg.org/ resources) (Zhou et al., 2018).”
A resource-efficient tool for mixed model association analysis of large-scale data (2019, methodology): “…including BOLT-LMM because although in theory MLM-based methods accounts for population stratification by fitting all (or a subset of selected) variants as random effects 29,30,39 , MLM-based association analyses in large samples suggest that fitting PCs as covariates improves robustness 42,43 .”