Sample Relatedness Research Articles

Inflation of polygenic risk scores caused by sample overlap and relatedness: Examples of a major risk of bias

Polygenic risk scores (PRSs) are an important tool for understanding the role of common genetic variants in human disease. Standard best practices recommend that PRSs be analyzed in cohorts that are independent of the genome-wide association study (GWAS) used to derive the scores without sample overlap or relatedness between the two cohorts. However, identifying sample overlap and relatedness can be challenging in an era of GWASs performed by large biobanks and international research consortia. Although most genomics researchers are aware of best practices and theoretical concerns about sample overlap and relatedness between GWAS and PRS cohorts, the prevailing assumption is that the risk of bias is small for very large GWASs. Here, we present two real-world examples demonstrating that sample overlap and relatedness is not a minor or theoretical concern but an important potential source of bias in PRS studies. Using a recently developed statistical adjustment tool, we found that excluding overlapping and related samples was equal to or more powerful than adjusting for overlap bias. Our goal is to make genomics researchers aware of the magnitude of risk of bias from sample overlap and relatedness and to highlight the need for mitigation tools, including independent validation cohorts in PRS studies, continued development of statistical adjustment methods, and tools for researchers to test their cohorts for overlap and relatedness with GWAS cohorts without sharing individual-level data.

Detection of lumpy skin disease virus reads in the human upper respiratory tract microbiome requires further investigation.

Lumpy skin disease virus (LSDV), a double-stranded DNA virus from the Capripoxvirus genus, primarily affects Bos indicus, Bos taurus breeds, and water buffalo. Arthropod vectors, including mosquitoes and biting flies, are the main LSDV transmitters. Although LSDV is not zoonotic, this study unexpectedly detected LSDV reads in the upper respiratory tract microbiome of humans from rural and urban areas in Maharashtra, India. Nasopharyngeal and oropharyngeal swab samples collected for SARS-CoV-2 surveillance underwent whole-genome metagenomics sequencing, revealing LSDV reads in 25% of samples. Split kmer analysis provided insights into sample relatedness despite the low coverage of LSDV reads with the reference genome. Our findings, which include the detection of LSDV contigs aligning to specific locations on the reference genome, suggest a common source for LSDV reads, potentially shared water sources, or milk/milk products. Further investigation is needed to ascertain the mode of transmission and reason for the detection of LSDV reads in human upper respiratory tract.

Bayesian mixed model inference for genetic association under related samples with brain network phenotype.

Genetic association studies for brain connectivity phenotypes have gained prominence due to advances in noninvasive imaging techniques and quantitative genetics. Brain connectivity traits, characterized by network configurations and unique biological structures, present distinct challenges compared to other quantitative phenotypes. Furthermore, the presence of sample relatedness in the most imaging genetics studies limits the feasibility of adopting existing network-response modeling. In this article, we fill this gap by proposing a Bayesian network-response mixed-effect model that considers a network-variate phenotype and incorporates population structures including pedigrees and unknown sample relatedness. To accommodate the inherent topological architecture associated with the genetic contributions to the phenotype, we model the effect components via a set of effect network configurations and impose an inter-network sparsity and intra-network shrinkage to dissect the phenotypic network configurations affected by the risk genetic variant. A Markov chain Monte Carlo (MCMC) algorithm is further developed to facilitate uncertainty quantification. We evaluate the performance of our model through extensive simulations. By further applying the method to study, the genetic bases for brain structural connectivity using data from the Human Connectome Project with excessive family structures, we obtain plausible and interpretable results. Beyond brain connectivity genetic studies, our proposed model also provides a general linear mixed-effect regression framework for network-variate outcomes.

CSF metabolites associated with biomarkers of Alzheimer's disease pathology.

Metabolomics technology facilitates studying associations between small molecules and disease processes. Correlating metabolites in cerebrospinal fluid (CSF) with Alzheimer's disease (AD) CSF biomarkers may elucidate additional changes that are associated with early AD pathology and enhance our knowledge of the disease. The relative abundance of untargeted metabolites was assessed in 161 individuals from the Wisconsin Registry for Alzheimer's Prevention. A metabolome-wide association study (MWAS) was conducted between 269 CSF metabolites and protein biomarkers reflecting brain amyloidosis, tau pathology, neuronal and synaptic degeneration, and astrocyte or microglial activation and neuroinflammation. Linear mixed-effects regression analyses were performed with random intercepts for sample relatedness and repeated measurements and fixed effects for age, sex, and years of education. The metabolome-wide significance was determined by a false discovery rate threshold of 0.05. The significant metabolites were replicated in 154 independent individuals from then Wisconsin Alzheimer's Disease Research Center. Mendelian randomization was performed using genome-wide significant single nucleotide polymorphisms from a CSF metabolites genome-wide association study. Metabolome-wide association study results showed several significantly associated metabolites for all the biomarkers except Aβ42/40 and IL-6. Genetic variants associated with metabolites and Mendelian randomization analysis provided evidence for a causal association of metabolites for soluble triggering receptor expressed on myeloid cells 2 (sTREM2), amyloid β (Aβ40), α-synuclein, total tau, phosphorylated tau, and neurogranin, for example, palmitoyl sphingomyelin (d18:1/16:0) for sTREM2, and erythritol for Aβ40 and α-synuclein. This study provides evidence that CSF metabolites are associated with AD-related pathology, and many of these associations may be causal.

Scalable mixed model methods for set-based association studies on large-scale categorical data analysis and its application to exome-sequencing data in UK Biobank

The ongoing release of large-scale sequencing data in the UK Biobank allows for the identification of associations between rare variants and complex traits. SAIGE-GENE+ is a valid approach to conducting set-based association tests for quantitative and binary traits. However, for ordinal categorical phenotypes, applying SAIGE-GENE+ with treating the trait as quantitative or binarizing the trait can cause inflated type I error rates or power loss. In this study, we propose a scalable and accurate method for rare-variant association tests, POLMM-GENE, in which we used a proportional odds logistic mixed model to characterize ordinal categorical phenotypes while adjusting for sample relatedness. POLMM-GENE fully utilizes the categorical nature of phenotypes and thus can well control type I error rates while remaining powerful. In the analyses of UK Biobank 450k whole-exome-sequencing data for five ordinal categorical traits, POLMM-GENE identified 54 gene-phenotype associations.

Metagenomic Antimicrobial Susceptibility Testing from Simulated Native Patient Samples

Genomic antimicrobial susceptibility testing (AST) has been shown to be accurate for many pathogens and antimicrobials. However, these methods have not been systematically evaluated for clinical metagenomic data. We investigate the performance of in-silico AST from clinical metagenomes (MG-AST). Using isolate sequencing data from a multi-center study on antimicrobial resistance (AMR) as well as shotgun-sequenced septic urine samples, we simulate over 2000 complicated urinary tract infection (cUTI) metagenomes with known resistance phenotype to 5 antimicrobials. Applying rule-based and machine learning-based genomic AST classifiers, we explore the impact of sequencing depth and technology, metagenome complexity, and bioinformatics processing approaches on AST accuracy. By using an optimized metagenomics assembly and binning workflow, MG-AST achieved balanced accuracy within 5.1% of isolate-derived genomic AST. For poly-microbial infections, taxonomic sample complexity and relatedness of taxa in the sample is a key factor influencing metagenomic binning and downstream MG-AST accuracy. We show that the reassignment of putative plasmid contigs by their predicted host range and investigation of whole resistome capabilities improved MG-AST performance on poly-microbial samples. We further demonstrate that machine learning-based methods enable MG-AST with superior accuracy compared to rule-based approaches on simulated native patient samples.

EraSOR: a software tool to eliminate inflation caused by sample overlap in polygenic score analyses.

Polygenic risk score (PRS) analyses are now routinely applied across biomedical research. However, as PRS studies grow in size, there is an increased risk of sample overlap between the genome-wide association study (GWAS) from which the PRS is derived and the "target sample," in which PRSs are computed and hypotheses are tested. Despite the wide recognition of the sample overlap problem, its potential impact on the results from PRS studies has not yet been quantified, and no analytical solution has been provided. Here, we first conduct a comprehensive investigation into the scale of the sample overlap problem, finding that PRS results can be substantially inflated even in the presence of minimal overlap. Next, we introduce a method and software, EraSOR (Erase Sample Overlap and Relatedness), which eliminates the inflation caused by sample overlap (and close relatedness) in almost all settings tested here. EraSOR could be useful in PRS studies (with target sample >1,000) similar to those investigated here, either (i) to mitigate the potential effects of known or unknown intercohort overlap and close relatedness or (ii) as a sensitivity tool to highlight the possible presence of sample overlap before its direct removal, when possible, or else to provide a lower bound on PRS analysis results after accounting for potential sample overlap.

A fast and powerful linear mixed model approach for genotype-environment interaction tests in large-scale GWAS.

Genotype-by-environment interaction (GEI or GxE) plays an important role in understanding complex human traits. However, it is usually challenging to detect GEI signals efficiently and accurately while adjusting for population stratification and sample relatedness in large-scale genome-wide association studies (GWAS). Here we propose a fast and powerful linear mixed model-based approach, fastGWA-GE, to test for GEI effect and G + GxE joint effect. Our extensive simulations show that fastGWA-GE outperforms other existing GEI test methods by controlling genomic inflation better, providing larger power and running hundreds to thousands of times faster. We performed a fastGWA-GE analysis of ~7.27 million variants on 452 249 individuals of European ancestry for 13 quantitative traits and five environment variables in the UK Biobank GWAS data and identified 96 significant signals (72 variants across 57 loci) with GEI test P-values < 1 × 10-9, including 27 novel GEI associations, which highlights the effectiveness of fastGWA-GE in GEI signal discovery in large-scale GWAS.

Quality Control Procedures for Genome-Wide Association Studies.

Genome-wide association studies (GWAS) are being conducted at an unprecedented rate in population-based cohorts and have increased our understanding of the pathophysiology of many complex diseases. Regardless of the context, the practical utility of this information ultimately depends upon the quality of the data used for statistical analyses. Quality control (QC) procedures for GWAS are constantly evolving. Here, we enumerate some of the challenges in QC of genotyped GWAS data and describe the approaches involving genotype imputation of a sample dataset along with post-imputation quality assurance, thereby minimizing potential bias and error in GWAS results. We discuss common issues associated with QC of the GWAS data (genotyped and imputed), including data file formats, software packages for data manipulation and analysis, sex chromosome anomalies, sample identity, sample relatedness, population substructure, batch effects, and marker quality. We provide detailed guidelines along with a sample dataset to suggest current best practices and discuss areas of ongoing and future research. © 2022 Wiley Periodicals LLC.

Extensive simulations assess the performance of genome-wide association mapping in variousSaccharomyces cerevisiaesubpopulations

With the advent of high throughput sequencing technologies, genome-wide association studies (GWAS) have become a powerful paradigm for dissecting the genetic origins of the observed phenotypic variation. We recently completely sequenced the genome of 1011 Saccharomyces cerevisiae isolates, laying a strong foundation for GWAS. To assess the feasibility and the limits of this approach, we performed extensive simulations using five selected subpopulations as well as the total set of 1011 genomes. We measured the ability to detect the causal genetic variants involved in Mendelian and more complex traits using a linear mixed model approach. The results showed that population structure is well accounted for and is not the main problem when the sample size is high enough. While the genetic determinant of a Mendelian trait is easily mapped in all studied subpopulations, discrepancies are seen between datasets when performing GWAS on a complex trait in terms of detection, false positive and false negative rate. Finally, we performed GWAS on the different defined subpopulations using a real quantitative trait (resistance to copper sulfate) and showed the feasibility of this approach. The performance of each dataset depends simultaneously on several factors such as sample size, relatedness and population evolutionary history.This article is part of the theme issue ‘Genetic basis of adaptation and speciation: from loci to causative mutations’.

Computationally efficient whole-genome regression for quantitative and binary traits.

Genome-wide association analysis of cohorts with thousands of phenotypes is computationally expensive, particularly when accounting for sample relatedness or population structure. Here we present a novel machine-learning method called REGENIE for fitting a whole-genome regression model for quantitative and binary phenotypes that is substantially faster than alternatives in multi-trait analyses while maintaining statistical efficiency. The method naturally accommodates parallel analysis of multiple phenotypes and requires only local segments of the genotype matrix to be loaded in memory, in contrast to existing alternatives, which must load genome-wide matrices into memory. This results in substantial savings in compute time and memory usage. We introduce a fast, approximate Firth logistic regression test for unbalanced case-control phenotypes. The method is ideally suited to take advantage of distributed computing frameworks. We demonstrate the accuracy and computational benefits of this approach using the UK Biobank dataset with up to 407,746 individuals.

Variant-set association test for generalized linear mixed model.

Advances in high-throughput biotechnologies have culminated in a wide range of omics (such as genomics, epigenomics, transcriptomics, metabolomics, and metagenomics) studies, and increasing evidence in these studies indicates that the biological architecture of complex traits involves a large number of omics variants each with minor effects but collectively accounting for the full phenotypic variability. Thus, a major challenge in many "ome-wide" association analyses is to achieve adequate statistical power to identify multiple variants of small effect sizes, which is notoriously difficult for studies with relatively small-sample sizes. A small-sample adjustment incorporated in the kernel machine regression framework was proposed to solve this for association studies under various settings. However, such an adjustment in the generalized linear mixed model (GLMM) framework, which accounts for both sample relatedness and non-Gaussian outcomes, has not yet been attempted. In this study, we fill this gap by extending small-sample adjustment in kernel machine association test to GLMM. We propose a new Variant-Set Association Test (VSAT), a powerful and efficient analysis tool in GLMM, to examine the association between a set of omics variants and correlated phenotypes. The usefulness of VSAT is demonstrated using both numerical simulation studies and applications to data collected from multiple association studies. The software for implementing the proposed method in R is available at https://www.github.com/jchen1981/SSKAT.

301 Methods of genome-wide association studies and their applications in dairy cattle

Abstract Genome-wide association studies (GWAS) has been widely used to map quantitative trait loci (QTL) of complex traits and diseases since 2007. To date, the human GWAS catalog has accumulated 4,410 publications and 172,351 associations, and the animal QTLdb has curated 983 publications and 130,407 QTLs for cattle, largest in livestock species. During the past 13 years of development, GWAS methods has evolved from simple linear regression, using principal components to address sample relatedness, mixed models, to Bayesian full model approaches. These methods have their advantages and limitations, so it is important to choose an appropriate method, especially for studies in livestock where sample size is often limited. Note that the most popular GWAS approach, the mixed model method, originated from animal breeding and genetics research. Leveraging the national cattle genomic database at the Council on Dairy Cattle Breeding (CDCB), we have conducted GWAS analyses of various dairy traits to identify QTLs and SNP markers of importance. Combining with sequence and functional annotation data, we seek to understand the genetic basis of complex traits and to reveal useful knowledge that can be incorporated into more accurate genomic predictions in the future.

SAIGEgds-an efficient statistical tool for large-scale PheWAS with mixed models.

Phenome-wide association studies (PheWASs) are known to be a powerful tool in discovery and replication of genetic association studies. To reduce the computational burden of PheWAS in the large cohorts, such as the UK Biobank, the SAIGE method has been proposed to control for case-control imbalance and sample relatedness in a tractable manner. However, SAIGE is still computationally intensive when deployed in analyzing the associations of thousands of ICD10-coded phenotypes with whole-genome imputed genotype data. Here, we present a new high-performance statistical R package (SAIGEgds) for large-scale PheWAS using generalized linear mixed models. The package implements the SAIGE method in optimized C++ codes, taking advantage of sparse genotype dosages and integrating the efficient genomic data structure file format. Benchmarks using the UK Biobank White British genotype data (N ≈ 430K) with coronary heart disease and simulated cases show that the implementation in SAIGEgds is 5-6 times faster than the SAIGE R package. When used in conjunction with high-performance computing clusters, SAIGEgds provides an efficient analysis pipeline for biobank-scale PheWAS. https://bioconductor.org/packages/SAIGEgds; vignettes included. Supplementary data are available at Bioinformatics online.

Scalable generalized linear mixed model for region-based association tests in large biobanks and cohorts.

With very large sample sizes, biobanks provide an exciting opportunity to identify genetic components of complex traits. To analyze rare variants, region-based multiple-variant aggregate tests are commonly used to increase power for association tests. However, because of the substantial computational cost, existing region-based tests cannot analyze hundreds of thousands of samples while accounting for confounders such as population stratification and sample relatedness. Here we propose a scalable generalized mixed-model region-based association test, SAIGE-GENE, that is applicable to exome-wide and genome-wide region-based analysis for hundreds of thousands of samples and can account for unbalanced case-control ratios for binary traits. Through extensive simulation studies and analysis of the HUNT study with 69,716 Norwegian samples and the UK Biobank data with 408,910 White British samples, we show that SAIGE-GENE can efficiently analyze large-sample data (N > 400,000) with type I error rates well controlled.

A unified method for rare variant analysis of gene-environment interactions.

Advanced technology in whole-genome sequencing has offered the opportunity to comprehensively investigate the genetic contribution, particularly rare variants, to complex traits. Several region-based tests have been developed to jointly model the marginal effect of rare variants, but methods to detect gene-environment (GE) interactions are underdeveloped. Identifying the modification effects of environmental factors on genetic risk poses a considerable challenge. To tackle this challenge, we develop a method to detect GE interactions for rare variants using generalized linear mixed effect model. The proposed method can accommodate either binary or continuous traits in related or unrelated samples. Under this model, genetic main effects, GE interactions, and sample relatedness are modeled as random effects. We adopt a kernel-based method to leverage the joint information across rare variants and implement variance component score tests to reduce the computational burden. Our simulation studies of continuous and binary traits show that the proposed method maintains correct type I error rates and appropriate power under various scenarios, such as genotype main effects and GE interaction effects in opposite directions and varying the proportion of causal variants in the model. We apply our method in the Framingham Heart Study to test GE interaction of smoking on body mass index or overweight status and replicate the Cholinergic Receptor Nicotinic Beta 4 gene association reported in previous large consortium meta-analysis of single nucleotide polymorphism-smoking interaction. Our proposed set-based GE test is computationally efficient and is applicable to both binary and continuous phenotypes, while appropriately accounting for familial or cryptic relatedness.

PolyQTL: Bayesian multiple eQTL detection with control for population structure and sample relatedness.

Expression quantitative loci (eQTL) are being used widely to annotate and interpret GWAS hits. Recent studies have demonstrated that individual gene expression is often regulated by multiple independent cis-acting eQTL. Diverse methods, frequentist and Bayesian, have already been developed to simultaneously detect and fine-map such multiple eQTL, but most of these ignore sample relatedness and potential population structure. This can result in false positives and disrupt the accuracy of fine-mapping. Here we introduce PolyQTL software for identifying and estimating eQTL effects. The package incorporates a genetic relatedness matrix to remove the influence of population structure and sample relatedness, while utilizing a Bayesian multiple eQTL detection pipeline to identify the most plausible candidate causal variants at one or more independent loci influencing abundance of a transcript. Simulations demonstrate that our approach improves the rate of discovery of causal variants relative to methods that do not account for relatedness. The software is written in C++, and freely available for download at https://github.com/jxzb1988/PolyQTL.

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

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, linear mixed model and the recently proposed logistic mixed model, perform poorly - producing large type I error rates - in the analysis of 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, provides accurate p-values even when case-control ratios are extremely unbalanced. It utilizes state-of-art optimization strategies to reduce computational cost, and hence is applicable to GWAS for thousands of phenotypes by large biobanks. Through the analysis of UK Biobank data of 408,961 white British European-ancestry samples for >1400 binary phenotypes, we show that SAIGE can efficiently analyze large sample data, controlling for unbalanced case-control ratios and sample relatedness.

A Bayesian Framework for Generalized Linear Mixed Modeling Identifies New Candidate Loci for Late-Onset Alzheimer's Disease.

Recent technical and methodological advances have greatly enhanced genome-wide association studies (GWAS). The advent of low-cost, whole-genome sequencing facilitates high-resolution variant identification, and the development of linear mixed models (LMM) allows improved identification of putatively causal variants. While essential for correcting false positive associations due to sample relatedness and population stratification, LMMs have commonly been restricted to quantitative variables. However, phenotypic traits in association studies are often categorical, coded as binary case-control or ordered variables describing disease stages. To address these issues, we have devised a method for genomic association studies that implements a generalized LMM (GLMM) in a Bayesian framework, called Bayes-GLMM. Bayes-GLMM has four major features: (1) support of categorical, binary, and quantitative variables; (2) cohesive integration of previous GWAS results for related traits; (3) correction for sample relatedness by mixed modeling; and (4) model estimation by both Markov chain Monte Carlo sampling and maximal likelihood estimation. We applied Bayes-GLMM to the whole-genome sequencing cohort of the Alzheimer’s Disease Sequencing Project. This study contains 570 individuals from 111 families, each with Alzheimer’s disease diagnosed at one of four confidence levels. Using Bayes-GLMM we identified four variants in three loci significantly associated with Alzheimer’s disease. Two variants, rs140233081 and rs149372995, lie between PRKAR1B and PDGFA. The coded proteins are localized to the glial-vascular unit, and PDGFA transcript levels are associated with Alzheimer’s disease-related neuropathology. In summary, this work provides implementation of a flexible, generalized mixed-model approach in a Bayesian framework for association studies.

NOVEL CANDIDATE LOCI FOR LATE-ONSET ALZHEIMER’S DISEASE FROM BAYESIAN MIXED MODELING OF WHOLE-GENOME AND WHOLE-EXOME SEQUENCING

Recent initiatives such as the Alzheimer's Disease Sequencing Project (ADSP) have created sequencing-based study cohorts for late-onset Alzheimer's disease. While this resource provides nucleotide-level resolution of genetic variants, it suffers from relatively small sample size and population substructure that limits association power. We addressed this limitation by devising a method of genome-wide association that implements a generalized linear mixed model in a Bayesian context. Our approach, Bayes-GLMM, has four main features: (1) support of categorical, binary and quantitative variables; (2) cohesive integration of previous GWAS results for related traits; (3) correction for sample relatedness by mixed modeling; and (4) model estimation by both Markov chain Monte Carlo (MCMC) sampling and maximal likelihood estimation. We applied Bayes-GLMM to the ADSP whole-genome sequencing cohort, which includes categorical disease diagnoses of 570 individuals drawn from 111 families. We also analyzed ADSP whole-exome data from 9134 individuals to identify rare and common coding variants. From the whole-genome sequencing cohort, we identified four variants in three loci significantly associated with Alzheimer's disease. The loci were not identified using traditional methods. The four variants (rs10490263, rs74944275, rs149372995, rs140233081) are located in intergenic regions with the closest genes not previously associated with AD. The two linked variants lie between the genes PRKAR1B and PDGFA. These proteins are localized to the glial-vascular unit, further implicating vascular function in modifying susceptibility to AD. Analysis of whole-exome sequences validated association of many known loci (e.g. TREM2, ABCA7) and identified potentially novel variants, including both rare and common alleles. This work provides new candidates loci for Alzheimer's disease, obtained from the first implementation of a flexible, generalized mixed model approach in a Bayesian framework for sequence-based association studies.