Abstract:Joint genetic models for multiple traits have helped to enhance association analyses. Most existing multi-trait models have been designed to increase power for detecting associations, whereas the analysis of interactions has received considerably less attention. Here, we propose iSet, a method based on linear mixed models to test for interactions between sets of variants and environmental states or other contexts. Our model generalizes previous interaction tests and in particular provides a test for local diff… Show more
“…1 b). Since it has previously been shown that inconsistent directions of effect for eQTLs will often arise from allelic heterogeneity rather than true sharing [ 20 , 21 ], we constrained factors to be nonnegative. Fig.…”
Section: Resultsmentioning
confidence: 99%
“…This discrepancy arose due to collinearity between the factors, and in such cases, the discrepant factors were not included for downstream analysis. We also removed those factors that caused one tissue to have an oppositely signed small effect (absolute Z -score < 3, or P value > 0.00135) when compared to the factor where this eQTL has the strongest effect; such discrepancies may often reflect allelic heterogeneity or LD contamination rather than true opposite effects from the same causal variant [ 20 , 21 ]…”
Section: Methodsmentioning
confidence: 99%
“…This discrepancy arose due to collinearity between the factors, and in such cases, the discrepant factors were not included for downstream analysis. We also removed those factors that caused one tissue to have an oppositely signed small effect (absolute Z-score < 3, or P value > 0.00135) when compared to the factor where this eQTL has the strongest effect; such discrepancies may often reflect allelic heterogeneity or LD contamination rather than true opposite effects from the same causal variant [20,21] Background SNP-gene pairs For enrichment analyses, random SNP-gene pairs were sampled from all SNP-gene pairs to match for eQTLs by three criteria: (1) SNP MAF was matched to the eQTL variants' MAF, (2) distance from the SNP to transcription start sites (TSS) of the gene was matched to eQTL, and (3) a number of SNPs per gene were matched as in eQTLs.…”
Genetic regulation of gene expression, revealed by expression quantitative trait loci (eQTLs), exhibits complex patterns of tissue-specific effects. Characterization of these patterns may allow us to better understand mechanisms of gene regulation and disease etiology. We develop a constrained matrix factorization model, sn-spMF, to learn patterns of tissue-sharing and apply it to 49 human tissues from the Genotype-Tissue Expression (GTEx) project. The learned factors reflect tissues with known biological similarity and identify transcription factors that may mediate tissue-specific effects. sn-spMF, available at https://github.com/heyuan7676/ts_eQTLs, can be applied to learn biologically interpretable patterns of eQTL tissue-specificity and generate testable mechanistic hypotheses.
“…1 b). Since it has previously been shown that inconsistent directions of effect for eQTLs will often arise from allelic heterogeneity rather than true sharing [ 20 , 21 ], we constrained factors to be nonnegative. Fig.…”
Section: Resultsmentioning
confidence: 99%
“…This discrepancy arose due to collinearity between the factors, and in such cases, the discrepant factors were not included for downstream analysis. We also removed those factors that caused one tissue to have an oppositely signed small effect (absolute Z -score < 3, or P value > 0.00135) when compared to the factor where this eQTL has the strongest effect; such discrepancies may often reflect allelic heterogeneity or LD contamination rather than true opposite effects from the same causal variant [ 20 , 21 ]…”
Section: Methodsmentioning
confidence: 99%
“…This discrepancy arose due to collinearity between the factors, and in such cases, the discrepant factors were not included for downstream analysis. We also removed those factors that caused one tissue to have an oppositely signed small effect (absolute Z-score < 3, or P value > 0.00135) when compared to the factor where this eQTL has the strongest effect; such discrepancies may often reflect allelic heterogeneity or LD contamination rather than true opposite effects from the same causal variant [20,21] Background SNP-gene pairs For enrichment analyses, random SNP-gene pairs were sampled from all SNP-gene pairs to match for eQTLs by three criteria: (1) SNP MAF was matched to the eQTL variants' MAF, (2) distance from the SNP to transcription start sites (TSS) of the gene was matched to eQTL, and (3) a number of SNPs per gene were matched as in eQTLs.…”
Genetic regulation of gene expression, revealed by expression quantitative trait loci (eQTLs), exhibits complex patterns of tissue-specific effects. Characterization of these patterns may allow us to better understand mechanisms of gene regulation and disease etiology. We develop a constrained matrix factorization model, sn-spMF, to learn patterns of tissue-sharing and apply it to 49 human tissues from the Genotype-Tissue Expression (GTEx) project. The learned factors reflect tissues with known biological similarity and identify transcription factors that may mediate tissue-specific effects. sn-spMF, available at https://github.com/heyuan7676/ts_eQTLs, can be applied to learn biologically interpretable patterns of eQTL tissue-specificity and generate testable mechanistic hypotheses.
“…1a). Recent extensions enable assessing GxE across sets of genetic variants, either using genetic risk scores6 or variance component tests7–9. Whilst there is evidence that multiple environments can interact with a single genetic locus to influence phenotypes, for example a number of environments have been shown to alter the effect of FTO on BMI, including physical activity10–13, diet12–15 and smoking12, there are no robust methods for the joint GxE analysis of multiple environmental variables.…”
Different exposures, including diet, physical activity, or external
conditions can contribute to genotype-environment interactions (GxE). Although
high-dimensional environmental data are increasingly available, and multiple
exposures have been implicated with GxE at the same loci, multi-environment
tests for GxE are not established. Here, we propose the structured linear mixed
model (StructLMM), a computationally efficient method to identify and
characterize loci that interact with one or more environments. After validating
our model using simulations, we apply StructLMM to body mass index in UK
Biobank, where our model yields previously known and novel GxE signals. Finally,
in an application to a large blood eQTL dataset, we demonstrate that StructLMM
can be used to study interactions with hundreds of environmental variables.
“…As environmental data are increasingly common in largescale human genetic studies, interaction analyses including multiple SNPs and multiple exposures might be performed systematically on behalf of standard G-E interaction screenings. However, despite a few recent works published (Casale et al 2017), our knowledge of the strengths and limitations of joint analysis approach for multiple G-E interactions is still limited. Here, we addressed a part of this question and explored the relative performance of four joint G-E interaction test approaches for both quantitative and binary trait models:…”
With growing human genetic and epidemiologic data, there has been increased interest for the study of gene-byenvironment (G-E) interaction effects. Still, major questions remain on how to test jointly a large number of interactions between multiple SNPs and multiple exposures. In this study, we first compared the relative performance of four fixed-effect joint analysis approaches using simulated data, considering up to 10 exposures and 300 SNPs: (1) omnibus test, (2) multi-exposure and genetic risk score (GRS) test, (3) multi-SNP and environmental risk score (ERS) test, and (4) GRS-ERS test. Our simulations explored both linear and logistic regression while considering three statistics: the Wald test, the Score test, and the likelihood ratio test (LRT). We further applied the approaches to three large sets of human cohort data (n = 37,664), focusing on type 2 diabetes (T2D), obesity, hypertension, and coronary heart disease with smoking, physical activity, diets, and total energy intake. Overall, GRS-based approaches were the most robust, and had the highest power, especially when the G-E interaction effects were correlated with the marginal genetic and environmental effects. We also observed severe miscalibration of joint statistics in logistic models when the number of events per variable was too low when using either the Wald test or LRT test. Finally, our real data application detected nominally significant interaction effects for three outcomes (T2D, obesity, and hypertension), mainly from the GRS-ERS approach. In conclusion, this study provides guidelines for testing multiple interaction parameters in modern human cohorts including extensive genetic and environmental data.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
