Limits...
Gene-based testing of interactions in association studies of quantitative traits.

Ma L, Clark AG, Keinan A - PLoS Genet. (2013)

Bottom Line: The framework of gene-based gene-gene interaction (GGG) tests combine marker-based interaction tests between all pairs of markers in two genes to produce a gene-level test for interaction between the two.Extensive simulations point to correct type I error rates of all tests and show that the two truncated tests are more powerful than the other tests in cases of markers involved in the underlying interaction not being directly genotyped and in cases of multiple underlying interactions.We conclude that our GGG tests show improved power to identify gene-level interactions in existing, as well as emerging, association studies.

View Article: PubMed Central - PubMed

Affiliation: Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, New York, USA. lm529@cornell.edu

ABSTRACT
Various methods have been developed for identifying gene-gene interactions in genome-wide association studies (GWAS). However, most methods focus on individual markers as the testing unit, and the large number of such tests drastically erodes statistical power. In this study, we propose novel interaction tests of quantitative traits that are gene-based and that confer advantage in both statistical power and biological interpretation. The framework of gene-based gene-gene interaction (GGG) tests combine marker-based interaction tests between all pairs of markers in two genes to produce a gene-level test for interaction between the two. The tests are based on an analytical formula we derive for the correlation between marker-based interaction tests due to linkage disequilibrium. We propose four GGG tests that extend the following P value combining methods: minimum P value, extended Simes procedure, truncated tail strength, and truncated P value product. Extensive simulations point to correct type I error rates of all tests and show that the two truncated tests are more powerful than the other tests in cases of markers involved in the underlying interaction not being directly genotyped and in cases of multiple underlying interactions. We applied our tests to pairs of genes that exhibit a protein-protein interaction to test for gene-level interactions underlying lipid levels using genotype data from the Atherosclerosis Risk in Communities study. We identified five novel interactions that are not evident from marker-based interaction testing and successfully replicated one of these interactions, between SMAD3 and NEDD9, in an independent sample from the Multi-Ethnic Study of Atherosclerosis. We conclude that our GGG tests show improved power to identify gene-level interactions in existing, as well as emerging, association studies.

Show MeSH

Related in: MedlinePlus

Graphical illustration of the framework of gene-based single-marker test and its generalization to a gene-based gene–gene interaction (GGG) test as proposed in this paper.While the former considers the P values of each single-marker test (A), a GGG test (B) is based on all P values of an interaction test between each pair of markers from each of the two genes. In order to combine these pairwise P values into a single test, a correlation matrix that concurrently accounts for linkage disequilibrium in each of the two genes needs to be estimated, which we derive in Materials and Methods.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3585009&req=5

pgen-1003321-g001: Graphical illustration of the framework of gene-based single-marker test and its generalization to a gene-based gene–gene interaction (GGG) test as proposed in this paper.While the former considers the P values of each single-marker test (A), a GGG test (B) is based on all P values of an interaction test between each pair of markers from each of the two genes. In order to combine these pairwise P values into a single test, a correlation matrix that concurrently accounts for linkage disequilibrium in each of the two genes needs to be estimated, which we derive in Materials and Methods.

Mentions: Gene-based tests of main association effects can be classified into two categories, tests that consider multiple markers in a gene as part of a joint model [39], [40], [41], [42], [43], [44], [45], [46] and tests that combine marker-based test statistics or P values into a gene-based equivalent (Figure 1A) [31], [32]. One important advantage of the latter type of tests, which are the focus of this paper, is that they do not require any additional information once the marker-based interaction P values have been evaluated. While it is imperative to account for the correlation between tests of different markers that is due to LD, this can be achieved using estimates from an external reference panel if genotype information is not available. Here, we propose four gene-based gene-gene interaction (GGG) tests of quantitative traits by extending four existing methods of combining P values: (i) minimum p value [32], (ii) extended Simes procedure (GATES) [31], (iii) truncated tail strength [47], and (iv) truncated-product P value [48]. Our tests employ these methods to combine P values of interaction tests between all pairs of individual SNPs to obtain a P value for a GGG test, while accounting for the correlation between the individual P values (Figure 1B). A recent study has recently extended ATOM [41], a gene-based main effect test of the type that considers all markers in a gene in a joint model, to a gene-based test that collapses all markers in each gene prior to interaction testing [14]. An advantage of the P value combining approaches is that if there are multiple heterogeneous interactions between a pair of genes, first collapsing SNPs in each gene according to the former approach can average out these disparate signals and lead to a reduction in power. Other than P value combining approaches, linkage disequilibrium has often been utilized for detection of gene-gene interactions in case-control studies. By comparing LD patterns between cases and controls, Rajapakse et al. have recently developed a gene-based test of interactions for case-control studies [26].


Gene-based testing of interactions in association studies of quantitative traits.

Ma L, Clark AG, Keinan A - PLoS Genet. (2013)

Graphical illustration of the framework of gene-based single-marker test and its generalization to a gene-based gene–gene interaction (GGG) test as proposed in this paper.While the former considers the P values of each single-marker test (A), a GGG test (B) is based on all P values of an interaction test between each pair of markers from each of the two genes. In order to combine these pairwise P values into a single test, a correlation matrix that concurrently accounts for linkage disequilibrium in each of the two genes needs to be estimated, which we derive in Materials and Methods.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3585009&req=5

pgen-1003321-g001: Graphical illustration of the framework of gene-based single-marker test and its generalization to a gene-based gene–gene interaction (GGG) test as proposed in this paper.While the former considers the P values of each single-marker test (A), a GGG test (B) is based on all P values of an interaction test between each pair of markers from each of the two genes. In order to combine these pairwise P values into a single test, a correlation matrix that concurrently accounts for linkage disequilibrium in each of the two genes needs to be estimated, which we derive in Materials and Methods.
Mentions: Gene-based tests of main association effects can be classified into two categories, tests that consider multiple markers in a gene as part of a joint model [39], [40], [41], [42], [43], [44], [45], [46] and tests that combine marker-based test statistics or P values into a gene-based equivalent (Figure 1A) [31], [32]. One important advantage of the latter type of tests, which are the focus of this paper, is that they do not require any additional information once the marker-based interaction P values have been evaluated. While it is imperative to account for the correlation between tests of different markers that is due to LD, this can be achieved using estimates from an external reference panel if genotype information is not available. Here, we propose four gene-based gene-gene interaction (GGG) tests of quantitative traits by extending four existing methods of combining P values: (i) minimum p value [32], (ii) extended Simes procedure (GATES) [31], (iii) truncated tail strength [47], and (iv) truncated-product P value [48]. Our tests employ these methods to combine P values of interaction tests between all pairs of individual SNPs to obtain a P value for a GGG test, while accounting for the correlation between the individual P values (Figure 1B). A recent study has recently extended ATOM [41], a gene-based main effect test of the type that considers all markers in a gene in a joint model, to a gene-based test that collapses all markers in each gene prior to interaction testing [14]. An advantage of the P value combining approaches is that if there are multiple heterogeneous interactions between a pair of genes, first collapsing SNPs in each gene according to the former approach can average out these disparate signals and lead to a reduction in power. Other than P value combining approaches, linkage disequilibrium has often been utilized for detection of gene-gene interactions in case-control studies. By comparing LD patterns between cases and controls, Rajapakse et al. have recently developed a gene-based test of interactions for case-control studies [26].

Bottom Line: The framework of gene-based gene-gene interaction (GGG) tests combine marker-based interaction tests between all pairs of markers in two genes to produce a gene-level test for interaction between the two.Extensive simulations point to correct type I error rates of all tests and show that the two truncated tests are more powerful than the other tests in cases of markers involved in the underlying interaction not being directly genotyped and in cases of multiple underlying interactions.We conclude that our GGG tests show improved power to identify gene-level interactions in existing, as well as emerging, association studies.

View Article: PubMed Central - PubMed

Affiliation: Department of Biological Statistics and Computational Biology, Cornell University, Ithaca, New York, USA. lm529@cornell.edu

ABSTRACT
Various methods have been developed for identifying gene-gene interactions in genome-wide association studies (GWAS). However, most methods focus on individual markers as the testing unit, and the large number of such tests drastically erodes statistical power. In this study, we propose novel interaction tests of quantitative traits that are gene-based and that confer advantage in both statistical power and biological interpretation. The framework of gene-based gene-gene interaction (GGG) tests combine marker-based interaction tests between all pairs of markers in two genes to produce a gene-level test for interaction between the two. The tests are based on an analytical formula we derive for the correlation between marker-based interaction tests due to linkage disequilibrium. We propose four GGG tests that extend the following P value combining methods: minimum P value, extended Simes procedure, truncated tail strength, and truncated P value product. Extensive simulations point to correct type I error rates of all tests and show that the two truncated tests are more powerful than the other tests in cases of markers involved in the underlying interaction not being directly genotyped and in cases of multiple underlying interactions. We applied our tests to pairs of genes that exhibit a protein-protein interaction to test for gene-level interactions underlying lipid levels using genotype data from the Atherosclerosis Risk in Communities study. We identified five novel interactions that are not evident from marker-based interaction testing and successfully replicated one of these interactions, between SMAD3 and NEDD9, in an independent sample from the Multi-Ethnic Study of Atherosclerosis. We conclude that our GGG tests show improved power to identify gene-level interactions in existing, as well as emerging, association studies.

Show MeSH
Related in: MedlinePlus