Genome-Wide Analysis of Gene-Gene and Gene-Environment Interactions Using Closed-Form Wald Tests

Causal Relationship Between Body Mass Index and Risk of Otitis Media with Effusion in Children: A Mendelian Randomization Study

Background

Body mass index(BMI) in children appears to be associated with Otitis media with effusion(OME) in observational studies, but the causal relationship is not clear.

Methods

A two-sample Mendelian randomization (MR) study was used to explore the causal relationship between childhood BMI and OME in people of European ancestry. Genome-wide association studies (GWAS) of childhood BMI were used as exposures (n = 61,111), while GWAS of OME were used as outcomes (n = 429,290). The weighted inverse variance method (IVW) was used as a baseline method to test for causality. In addition, MR-Egger, simple mode analysis, weighted median, and weighted mode were used as complementary methods.MR-PRESSO analysis, MR-Egger intercept analysis, and Cochran's Q statistical analysis were also used to detect possible directional heterogeneity and polymorphism. To assess this association, we used ratios (OR) with 95% confidence intervals (ci). All statistical analyses were performed in R.

Results

We selected 22 genome-wide significant single nucleotide polymorphisms (SNPs) from GWAS as instrumental variables (IVW). the IVW approach showed evidence supporting a causal relationship between BMI and OME in children (β = 0.265, SE = 0.113, P = 0.018). MR-Egger regression showed that targeted polymorphisms were unlikely to bias the results bias (intercept=-0.022; P = 0.488), but there was no causal relationship between BMI and OME (β = 0.584, SE = 0.465, P = 0.224). Although the results of the IVW and MR Egger analyses were not consistent, the IVW analysis maintained higher precision, and the Cochran Q test, heterogeneity and polymorphism tests showed no heterogeneity, no directionality and no polymorphism.

Conclusions

MR studies suggest that genetically predicted body mass index in childhood is associated with an increased risk of OME. Notably, given the limitations of this study, the mechanism of association between body mass index and OME in childhood needs further investigation. These results support the importance of effective management of obesity, which may reduce OME occurrence and decrease OME recurrence.

Metformin Treatment Reduces the Incidence of Rheumatoid Arthritis: A Two-Sample Mendelian Randomized Study

Several studies have shown that rheumatologic patients can benefit from metformin, but it remains unclear whether metformin treatment is causally associated with the risk of rheumatoid arthritis (RA). A two-sample Mendelian randomization (MR) study was conducted to investigate the causal relationship between metformin treatment and the incidence of rheumatoid arthritis. The genome-wide significant (p < 5 × 10^-8) single-nucleotide polymorphisms (SNPs) associated with metformin use were selected as instrumental variables (IVs). Summary statistics on RA were extracted from a large genome-wide association study (GWAS) meta-analysis. The inverse variance-weighted (IVW) method was used as the determinant of the causal effects of metformin treatment on RA. Cochran's Q was used to detect heterogeneity. Mendelian randomization pleiotropy residual sum and outlier (MR-PRESSO) test and MR-Egger regression were used to detect horizontal pleiotropy. A total of 34 SNPs significantly associated with metformin treatment were obtained. Thirty-two SNPs were selected as IVs after removing two SNPs for being palindromic with intermediate allele frequencies (rs11658063 and rs4930011). The IVW results showed a negative causal association between metformin treatment and RA (OR = 0.0232, 95% CI 1.6046 × 10^-3 - 0.3368; p = 0.006). Meanwhile, no heterogeneity or pleiotropy was detected, indicating that the results were reliable. This study indicated a negative causality between metformin treatment and RA, indicating that the treatment of metformin can prevent the pathogenesis of RA.

Inference of gene-environment interaction from heterogeneous case-parent trios

Introduction: In genetic epidemiology, log-linear models of population risk may be used to study the effect of genotypes and exposures on the relative risk of a disease. Such models may also include gene-environment interaction terms that allow the genotypes to modify the effect of the exposure, or equivalently, the exposure to modify the effect of genotypes on the relative risk. When a measured test locus is in linkage disequilibrium with an unmeasured causal locus, exposure-related genetic structure in the population can lead to spurious gene-environment interaction; that is, to apparent gene-environment interaction at the test locus in the absence of true gene-environment interaction at the causal locus. Exposure-related genetic structure occurs when the distributions of exposures and of haplotypes at the test and causal locus both differ across population strata. A case-parent trio design can protect inference of genetic main effects from confounding bias due to genetic structure in the population. Unfortunately, when the genetic structure is exposure-related, the protection against confounding bias for the genetic main effect does not extend to the gene-environment interaction term. Methods: We show that current methods to reduce the bias in estimated gene-environment interactions from case-parent trio data can only account for simple population structure involving two strata. To fill this gap, we propose to directly accommodate multiple population strata by adjusting for genetic principal components (PCs). Results and Discussion: Through simulations, we show that our PC adjustment maintains the nominal type-1 error rate and has nearly identical power to detect gene-environment interaction as an oracle approach based directly on population strata. We also apply the PC-adjustment approach to data from a study of genetic modifiers of cleft palate comprised primarily of case-parent trios of European and East Asian ancestry. Consistent with earlier analyses, our results suggest that the gene-environment interaction signal in these data is due to the self-reported European trios.

Estimating the causal effect of frailty index on vestibular disorders: A two-sample Mendelian randomization

Background

Frailty index and vestibular disorders appear to be associated in observational studies, but causality of the association remains unclear.

Methods

A two-sample Mendelian randomization (MR) study was implemented to explore the causal relationship between the frailty index and vestibular disorders in individuals of European descent. A genome-wide association study (GWAS) of frailty index was used as the exposure (n = 175, 226), whereas the GWAS of vestibular disorders was the outcome (n = 462,933). MR Steiger filtering method was conducted to investigate the causal effect of the frailty index on vestibular disorders. An inverse variance weighted (IVW) approach was used as the essential approach to examine the causality. Additionally, the MR-Egger methods, the simple mode analysis, the weighted median analysis, and the weighted mode analysis were used as supplementary methods. The MR-PRESSO analysis, the MR-Egger intercept analysis, and Cochran's Q statistical analysis also were used to detect the possible heterogeneity as well as directional pleiotropy. To evaluate this association, the odds ratio (OR) with 95% confidence intervals (CIs) was used. All statistical analyses were performed in R. The STROBE-MR checklist for the reporting of MR studies was used in this study.

Results

In total, 14 single nucleotide polymorphisms (SNPs) were identified as effective instrumental variables (IVs) in the two sample MR analyses. The significant causal effect of the frailty index on vestibular disorders was demonstrated by IVW method [OR 1.008 (95% CI 1.003, 1.013), p = 0.001]. Results from the various sensitivity analysis were consistent. The “leave-one-out” analysis indicated that our results were robust even without a single SNP. According to the MR-Egger intercept test [intercept = −0.000151, SE = 0.011, p = 0.544], genetic pleiotropy did not affect the results. No heterogeneity was detected by Cochran's Q test. Results of MR Steiger directionality test indicated the accuracy of our estimate of the potential causal direction (Steiger p < 0.001).

Conclusion

The MR study suggested that genetically predicted frailty index may be associated with an increased risk of vestibular disorders. Notably, considering the limitations of this study, the causal effects between frailty index and vestibular disorders need further investigation. These results support the importance of effectively managing frailty which may minimize vestibular disorders and improve the quality of life for those with vestibular disorders.

Is Hearing Impairment Causally Associated With Falls? Evidence From a Two-Sample Mendelian Randomization Study

Background

Observational studies have suggested that hearing impairment (HI) was associated with the risk of falls, but it remains unclear if this association is of causal nature.

Methods

A two-sample Mendelian randomization (MR) study was conducted to investigate the causal association between HI and falls in individuals of European descent. Summary data on the association of single nucleotide polymorphisms (SNPs) with HI were obtained from the hitherto largest genome-wide association study (GWAS) (n = 323,978), and statistics on the association of SNPs with falls were extracted from another recently published GWAS (n = 461,725). MR Steiger filtering method was applied to determine the causal direction between HI and falls. Inverse-variance weighted (IVW) method was employed as the main approach to analyze the causal association between HI and falls, whereas weighted median, simple mode, weighted mode, and MR-Egger methods were used as complementary analyses. The MR-Egger intercept test, the MR-PRESSO test, and Cochran's Q statistic were performed to detect the potential directional pleiotropy and heterogeneity, respectively. The odds ratio (OR) with 95% confidence intervals (CIs) was used to evaluate this association.

Results

A total of 18 SNPs were identified as valid instrumental variables in our two-sample MR analysis. The positive causality between HI and risk of falls was indicated by IVW [OR 1.108 (95% CI 1.028, 1.194), p = 0.007]. The sensitivity analyses yielded comparable results. The “leave-one-out” analysis proved that lack of a single SNP did not affect the robustness of our results. The MR-Egger intercept test exhibited that genetic pleiotropy did not bias the results [intercept = −2.4E−04, SE = 0.001, p = 0.832]. Cochran's Q test revealed no heterogeneity.

Conclusion

Our MR study revealed a causal association between genetically predicted HI and falls. These results provide further evidence supporting the need to effectively manage HI to minimize fall risks and improve quality of life.

A new method for exploring gene-gene and gene-environment interactions in GWAS with tree ensemble methods and SHAP values

Background

The identification of gene–gene and gene–environment interactions in genome-wide association studies is challenging due to the unknown nature of the interactions and the overwhelmingly large number of possible combinations. Parametric regression models are suitable to look for prespecified interactions. Nonparametric models such as tree ensemble models, with the ability to detect any unspecified interaction, have previously been difficult to interpret. However, with the development of methods for model explainability, it is now possible to interpret tree ensemble models efficiently and with a strong theoretical basis.

Results

We propose a tree ensemble- and SHAP-based method for identifying as well as interpreting potential gene–gene and gene–environment interactions on large-scale biobank data. A set of independent cross-validation runs are used to implicitly investigate the whole genome. We apply and evaluate the method using data from the UK Biobank with obesity as the phenotype. The results are in line with previous research on obesity as we identify top SNPs previously associated with obesity. We further demonstrate how to interpret and visualize interaction candidates.

Conclusions

The new method identifies interaction candidates otherwise not detected with parametric regression models. However, further research is needed to evaluate the uncertainties of these candidates. The method can be applied to large-scale biobanks with high-dimensional data.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-021-04041-7.

FGFR2 Extracellular Domain In-Frame Deletions are Therapeutically Targetable Genomic Alterations that Function as Oncogenic Drivers in Cholangiocarcinoma

We conducted next generation DNA sequencing on 335 biliary tract cancers and characterized the genomic landscape by anatomic site within the biliary tree. In addition to frequent FGFR2 fusions among patients with intrahepatic cholangiocarcinoma (IHCC), we identified FGFR2 extracellular domain in-frame deletions (EIDs) in 5 of 178 (2.8%) patients with IHCC, including two patients with FGFR2 p.H167_N173del. Expression of this FGFR2 EID in NIH3T3 cells resulted in constitutive FGFR2 activation, oncogenic transformation, and sensitivity to FGFR inhibitors. Three patients with FGFR2 EIDs were treated with Debio 1347, an oral FGFR-1/2/3 inhibitor, and all showed partial responses. One patient developed an acquired L618F FGFR2 kinase domain mutation at disease progression and experienced a further partial response for 17 months to an irreversible FGFR2 inhibitor, futibatinib. Together, these findings reveal FGFR2 EIDs as an alternative mechanism of FGFR2 activation in IHCC that predict sensitivity to FGFR inhibitors in the clinic.

Performance of epistasis detection methods in semi-simulated GWAS

BACKGROUND

Part of the missing heritability in Genome Wide Association Studies (GWAS) is expected to be explained by interactions between genetic variants, also called epistasis. Various statistical methods have been developed to detect epistasis in case-control GWAS. These methods face major statistical challenges due to the number of tests required, the complexity of the Linkage Disequilibrium (LD) structure, and the lack of consensus regarding the definition of epistasis. Their limited impact in terms of uncovering new biological knowledge might be explained in part by the limited amount of experimental data available to validate their statistical performances in a realistic GWAS context. In this paper, we introduce a simulation pipeline for generating real scale GWAS data, including epistasis and realistic LD structure. We evaluate five exhaustive bivariate interaction methods, fastepi, GBOOST, SHEsisEpi, DSS, and IndOR. Two hundred thirty four different disease scenarios are considered in extensive simulations. We report the performances of each method in terms of false positive rate control, power, area under the ROC curve (AUC), and computation time using a GPU. Finally we compare the result of each methods on a real GWAS of type 2 diabetes from the Welcome Trust Case Control Consortium.

RESULTS

GBOOST, SHEsisEpi and DSS allow a satisfactory control of the false positive rate. fastepi and IndOR present an increase in false positive rate in presence of LD between causal SNPs, with our definition of epistasis. DSS performs best in terms of power and AUC in most scenarios with no or weak LD between causal SNPs. All methods can exhaustively analyze a GWAS with 6.105 SNPs and 15,000 samples in a couple of hours using a GPU.

CONCLUSION

This study confirms that computation time is no longer a limiting factor for performing an exhaustive search of epistasis in large GWAS. For this task, using DSS on SNP pairs with limited LD seems to be a good strategy to achieve the best statistical performance. A combination approach using both DSS and GBOOST is supported by the simulation results and the analysis of the WTCCC dataset demonstrated that this approach can detect distinct genes in epistasis. Finally, weak epistasis between common variants will be detectable with existing methods when GWAS of a few tens of thousands cases and controls are available.

Fast and general tests of genetic interaction for genome-wide association studies

A complex disease has, by definition, multiple genetic causes. In theory, these causes could be identified individually, but their identification will likely benefit from informed use of anticipated interactions between causes. In addition, characterizing and understanding interactions must be considered key to revealing the etiology of any complex disease. Large-scale collaborative efforts are now paving the way for comprehensive studies of interaction. As a consequence, there is a need for methods with a computational efficiency sufficient for modern data sets as well as for improvements of statistical accuracy and power. Another issue is that, currently, the relation between different methods for interaction inference is in many cases not transparent, complicating the comparison and interpretation of results between different interaction studies. In this paper we present computationally efficient tests of interaction for the complete family of generalized linear models (GLMs). The tests can be applied for inference of single or multiple interaction parameters, but we show, by simulation, that jointly testing the full set of interaction parameters yields superior power and control of false positive rate. Based on these tests we also describe how to combine results from multiple independent studies of interaction in a meta-analysis. We investigate the impact of several assumptions commonly made when modeling interactions. We also show that, across the important class of models with a full set of interaction parameters, jointly testing the interaction parameters yields identical results. Further, we apply our method to genetic data for cardiovascular disease. This allowed us to identify a putative interaction involved in Lp(a) plasma levels between two ‘tag’ variants in the LPA locus (p = 2.42 ⋅ 10⁻⁰⁹) as well as replicate the interaction (p = 6.97 ⋅ 10⁻⁰⁷). Finally, our meta-analysis method is used in a small (N = 16,181) study of interactions in myocardial infarction.

Interaction between organic molecules forms the basis of all biological systems. The availability of high-throughput genotyping and sequencing platforms enables us to cost-effectively genotype a large number of individuals. For sufficiently large datasets it is possible to reconstruct the genetic dependencies that underlie complex traits and diseases. However, there is a need for efficient statistical methodologies that can tackle the large sample size and computational resources required to study interaction. In this work we provide theory that reduces the required computational resources, and enable multiple research groups to effectively combine their results.

Successful Replication of GWAS Hits for Multiple Sclerosis in 10,000 Germans Using the Exome Array

Genome-wide association studies (GWAS) successfully identified various chromosomal regions to be associated with multiple sclerosis (MS). The primary aim of this study was to replicate reported associations from GWAS using an exome array in a large German study. German MS cases (n = 4,476) and German controls (n = 5,714) were genotyped using the Illumina HumanExome v1-Chip. Genotype calling was performed with the Illumina Genome Studio(TM) Genotyping Module, followed by zCall. Single-nucleotide polymorphisms (SNPs) in seven regions outside the human leukocyte antigen (HLA) region showed genome-wide significant associations with MS (P values < 5 × 10(-8) ). These associations have been reported previously. In addition, SNPs in three previously reported regions outside the HLA region yielded P values < 10(-5) . The effect of nine SNPs in the HLA region remained (P < 10(-5) ) after adjustment for other significant SNPs in the HLA region. All of these findings have been reported before or are driven by known risk loci. In summary, findings from previous GWAS for MS could be successfully replicated. We conclude that the regions identified in previous GWAS are also associated in the German population. This reassures the need for detailed investigations of the functional mechanisms underlying the replicated associations.