Statistical analysis of genetic interactions

Gene-Gene Interaction Between Factor-XI and ABO Genes in Cerebral Venous Thrombosis: The BEAST Study

BACKGROUND AND OBJECTIVES

Gene-gene interactions likely contribute to the etiology of multifactorial diseases such as cerebral venous thrombosis (CVT) and could be one of the main sources of known missing heritability. We explored Factor XI (F11) and ABO gene interactions among patients with CVT.

METHODS

Patients with CVT of European ancestry from the large Bio-Repository to Establish the Aetiology of Sinovenous Thrombosis (BEAST) international collaboration were recruited. Codominant modelling was used to determine interactions between genome-wide identified F11 and ABO genes with CVT status.

RESULTS

We studied 882 patients with CVT and 1,205 ethnically matched control participants (age: 42 ± 15 vs 43 ± 12 years, p = 0.08: sex: 71% male vs 68% female, p = 0.09, respectively). Individuals heterozygous (AT) for the risk allele (T) at both loci (rs56810541/F11 and rs8176645/ABO) had a 3.9 (95% CI 2.74-5.71, p = 2.75e-13) increase in risk of CVT. Individuals homozygous (TT) for the risk allele at both loci had a 13.9 (95% CI 7.64-26.17, p = 2.0e-15) increase in risk of CVT. The presence of a non-O blood group (A, B, AB) combined with TT/rs56810541/F11 increased CVT risk by OR = 6.8 (95% CI 4.54-10.33, p = 2.00e15), compared with blood group-O combined with AA.

DISCUSSION

Interactions between factor XI and ABO genes increase risk of CVT by 4- to 14-fold.

A global genetic interaction network by single-cell imaging and machine learning

Cellular and organismal phenotypes are controlled by complex gene regulatory networks. However, reference maps of gene function are still scarce across different organisms. Here, we generated synthetic genetic interaction and cell morphology profiles of more than 6,800 genes in cultured Drosophila cells. The resulting map of genetic interactions was used for machine learning-based gene function discovery, assigning functions to genes in 47 modules. Furthermore, we devised Cytoclass as a method to dissect genetic interactions for discrete cell states at the single-cell resolution. This approach identified an interaction of Cdk2 and the Cop9 signalosome complex, triggering senescence-associated secretory phenotypes and immunogenic conversion in hemocytic cells. Together, our data constitute a genome-scale resource of functional gene profiles to uncover the mechanisms underlying genetic interactions and their plasticity at the single-cell level.

Impact of tooth mineral tissues genes on dental caries: a birth-cohort study

OBJECTIVE

We aimed to investigate whether Single Nucleotide Polymorphisms present in the genes of tooth mineral tissues influence dental caries trajectory across the life course, and if there is an epistatic (gene-gene) interaction between these SNPs.

METHODS

A representative sample of all 5,914 births from the 1982 Pelotas birth cohort study was prospectively investigated. Dental caries trajectory across the life course was assessed at 15(n=888), 24(n=720), and 31 years old(n=539). Group-based trajectory modelling was used to identify distinct subgroups of individuals whose caries measurements followed a similar pattern over time. Genetic material was collected, and individuals were genotyped [rs4970957(TUFT1), rs1711437(MMP20), rs1784418(MMP20), rs2252070(MMP13), rs243847(MMP2), rs2303466(DLX3), rs11656951(DLX3), rs7501477(TIMP2), rs388286(BMP7), and rs5997096(TFIP11)]. Analyzes were performed for allele and genotype using logistic regression and generalized multifactor dimensionality reduction for epistatic interactions.

RESULTS

The analyses included 678 individuals, those with allele C (OR=0.74, CI95%[0.59-0.92]), genotype CC in the additive effect (OR=0.52, CI95%[0.31-0.89]), and the genotype TC/CC in dominant effect (OR=0.72, CI95%[0.53-0.98]) on the rs243847(MMP2) were associated with low caries trajectory. Individuals with the allele T (OR=0.79, CI95%[0.64-0.98]) and the genotype TC/CC in dominant effect (OR=0.66, CI95%[0.47-0.95]) on the rs5997096(TFIP11) were associated with low caries trajectory. Positive epistatic interactions were observed involving two (MMP2 and BMP7; p=0.006) and three (TUFT1, MMP2, and TFIP11; p<0.001) loci and high caries trajectory.

CONCLUSIONS

Some SNPs present in the genes of tooth mineral tissues were associated with caries trajectory and epistatic interactions increasing the network of SNPs involved in individual caries experience.

CLINICAL SIGNIFICANCE

Single nucleotide polymorphisms in the pathway of tooth mineral tissues genes may contribute significantly to the individual caries experience across the life course.

Sweet Taste Receptor Gene and Caries Trajectory in the Life Course

This study aims to investigate whether the trajectory of dental caries in the life course is associated with rs307355 (TAS1R3) and rs35874116 (TAS1R2) and if there is an epistatic association between rs307355 (TAS1R3) and rs35874116 (TAS1R2). A representative sample of all 5,914 births from the 1982 Pelotas birth cohort was prospectively investigated, and the decayed, missing, and filled teeth (DMF-T) components were assessed at ages 15 (n = 888), 24 (n = 720), and 31 (n = 539) y. Group-based trajectory modeling was used to identify groups with similar trajectories of DMF-T components in the life course. Genetic material was collected, and rs307355 (TAS1R3) and rs35874116 (TAS1R2) were genotyped. Ethnicity was evaluated using ADMIXTURE. Generalized multifactor dimensionality reduction software was used to investigate epistatic interactions. Considering rs307355 (TAS1R3) in the additive effect, the genotype TT was associated with the high decayed trajectory group (odds ratio [OR] = 4.52; 95% confidence interval [CI], 1.15-17.74) and the high missing trajectory group (OR = 3.35; 95% CI, 1.09-10.26). In the dominant effect, the genotype CT/TT was associated with the high decayed trajectory group (OR = 1.64; 95% CI, 1.14-2.35). Allele T was associated with an increased odds of 64% (OR = 1.64; 95% CI, 1.20-2.25) for the decayed component and 41% (OR = 1.41; 95% CI, 1.04-1.92) for the missing component. No associations were observed between rs307355 (TAS1R3) and the filled component. rs35874116 (TAS1R2) was not associated with DMF-T components. Positive epistatic interactions were observed involving rs307355 (TAS1R3) and rs35874116 (TAS1R2) with the decayed component (OR = 1.72; 95% CI, 1.04-2.84). Thus, rs307355 (TAS1R3) genotypes and alleles seem positively associated with the trajectory of decayed and missing components in the life course. Epistatic interaction between rs307355 and rs35874116 may increase the decayed caries trajectory.

Machine Learning to Identify Interaction of Single-Nucleotide Polymorphisms as a Risk Factor for Chronic Drug-Induced Liver Injury

Drug-induced liver injury (DILI) is a major cause of drug development failure and drug withdrawal from the market after approval. The identification of human risk factors associated with susceptibility to DILI is of paramount importance. Increasing evidence suggests that genetic variants may lead to inter-individual differences in drug response; however, individual single-nucleotide polymorphisms (SNPs) usually have limited power to predict human phenotypes such as DILI. In this study, we aim to identify appropriate statistical methods to investigate gene-gene and/or gene-environment interactions that impact DILI susceptibility. Three machine learning approaches, including Multivariate Adaptive Regression Splines (MARS), Multifactor Dimensionality Reduction (MDR), and logistic regression, were used. The simulation study suggested that all three methods were robust and could identify the known SNP-SNP interaction when up to 4% of genotypes were randomly permutated. When applied to a real-life DILI chronicity dataset, both MARS and MDR, but not logistic regression, identified combined genetic variants having better associations with DILI chronicity in comparison to the use of individual SNPs. Furthermore, a simple decision tree model using the SNPs identified by MARS and MDR was developed to predict DILI chronicity, with fair performance. Our study suggests that machine learning approaches may help identify gene-gene interactions as potential risk factors for better assessing complicated diseases such as DILI chronicity.

Systems Approach to Pathogenic Mechanism of Type 2 Diabetes and Drug Discovery Design Based on Deep Learning and Drug Design Specifications

In this study, we proposed a systems biology approach to investigate the pathogenic mechanism for identifying significant biomarkers as drug targets and a systematic drug discovery strategy to design a potential multiple-molecule targeting drug for type 2 diabetes (T2D) treatment. We first integrated databases to construct the genome-wide genetic and epigenetic networks (GWGENs), which consist of protein–protein interaction networks (PPINs) and gene regulatory networks (GRNs) for T2D and non-T2D (health), respectively. Second, the relevant “real GWGENs” are identified by system identification and system order detection methods performed on the T2D and non-T2D RNA-seq data. To simplify network analysis, principal network projection (PNP) was thereby exploited to extract core GWGENs from real GWGENs. Then, with the help of KEGG pathway annotation, core signaling pathways were constructed to identify significant biomarkers. Furthermore, in order to discover potential drugs for the selected pathogenic biomarkers (i.e., drug targets) from the core signaling pathways, not only did we train a deep neural network (DNN)-based drug–target interaction (DTI) model to predict candidate drug’s binding with the identified biomarkers but also considered a set of design specifications, including drug regulation ability, toxicity, sensitivity, and side effects to sieve out promising drugs suitable for T2D.

Genome-wide association and epistatic interactions of flowering time in soybean cultivar

Genome-wide association studies (GWAS) have enabled the discovery of candidate markers that play significant roles in various complex traits in plants. Recently, with increased interest in the search for candidate markers, studies on epistatic interactions between single nucleotide polymorphism (SNP) markers have also increased, thus enabling the identification of more candidate markers along with GWAS on single-variant-additive-effect. Here, we focused on the identification of candidate markers associated with flowering time in soybean (Glycine max). A large population of 2,662 cultivated soybean accessions was genotyped using the 180k Axiom® SoyaSNP array, and the genomic architecture of these accessions was investigated to confirm the population structure. Then, GWAS was conducted to evaluate the association between SNP markers and flowering time. A total of 93 significant SNP markers were detected within 59 significant genes, including E1 and E3, which are the main determinants of flowering time. Based on the GWAS results, multilocus epistatic interactions were examined between the significant and non-significant SNP markers. Two significant and 16 non-significant SNP markers were discovered as candidate markers affecting flowering time via interactions with each other. These 18 candidate SNP markers mapped to 18 candidate genes including E1 and E3, and the 18 candidate genes were involved in six major flowering pathways. Although further biological validation is needed, our results provide additional information on the existing flowering time markers and present another option to marker-assisted breeding programs for regulating flowering time of soybean.

Suicidal ideation in systemic lupus erythematosus: NR2A gene polymorphism, clinical and psychosocial factors

Background Systemic lupus erythematosus (SLE) patients are a high-risk population for suicide. Glutamatergic neurosystem genes have been implicated in the neurobiology of depression in SLE and suicidal behaviour in general. However, the role of glutamate receptor gene polymorphisms in suicidal behaviour among SLE patients remains unclear in the context of established clinical and psychosocial factors. We aimed to investigate the association of NR2A gene polymorphism with suicidal ideation in SLE while accounting for the interaction between clinical and psychosocial factors. Methods A total of 130 SLE patients were assessed for mood disorders (MINI International Neuropsychiatric Interview), severity of depression (Patient Health Questionnaire-9), suicidal behaviour (Columbia-Suicide Severity Rating Scale), socio-occupational functioning (Work and Social Adjustment Scale), recent life events (Social Readjustment Rating Scale) and lupus disease activity (SELENA-SLE Disease Activity Index). Eighty-six out of the 130 study participants consented for NR2A genotyping. Results Multivariable logistic regression showed nominal significance for the interaction effect between the NR2A rs2072450 AC genotype and higher severity of socio-occupational impairment with lifetime suicidal ideation in SLE patients ( p = 0.038, odds ratio = 1.364, 95% confidence interval = 1.018-1.827). However, only the association between lifetime mood disorder and lifetime suicidal ideation remained significant after Bonferroni correction ( p < 0.001, odds ratio = 33.834, 95% confidence interval = 7.624-150.138). Conclusions Lifetime mood disorder emerged as a more significant factor for suicidal ideation in SLE compared with NR2A gene polymorphism main and interaction effects. Clinical implications include identification and treatment of mood disorders as an early intervention for suicidal behaviour in SLE. More adequately-powered gene-environment interaction studies are required in the future to clarify the role of glutamate receptor gene polymorphisms in the risk stratification of suicidal behaviour among SLE patients.

EM Adaptive LASSO-A Multilocus Modeling Strategy for Detecting SNPs Associated with Zero-inflated Count Phenotypes

Count data are increasingly ubiquitous in genetic association studies, where it is possible to observe excess zero counts as compared to what is expected based on standard assumptions. For instance, in rheumatology, data are usually collected in multiple joints within a person or multiple sub-regions of a joint, and it is not uncommon that the phenotypes contain enormous number of zeroes due to the presence of excessive zero counts in majority of patients. Most existing statistical methods assume that the count phenotypes follow one of these four distributions with appropriate dispersion-handling mechanisms: Poisson, Zero-inflated Poisson (ZIP), Negative Binomial, and Zero-inflated Negative Binomial (ZINB). However, little is known about their implications in genetic association studies. Also, there is a relative paucity of literature on their usefulness with respect to model misspecification and variable selection. In this article, we have investigated the performance of several state-of-the-art approaches for handling zero-inflated count data along with a novel penalized regression approach with an adaptive LASSO penalty, by simulating data under a variety of disease models and linkage disequilibrium patterns. By taking into account data-adaptive weights in the estimation procedure, the proposed method provides greater flexibility in multi-SNP modeling of zero-inflated count phenotypes. A fast coordinate descent algorithm nested within an EM (expectation-maximization) algorithm is implemented for estimating the model parameters and conducting variable selection simultaneously. Results show that the proposed method has optimal performance in the presence of multicollinearity, as measured by both prediction accuracy and empirical power, which is especially apparent as the sample size increases. Moreover, the Type I error rates become more or less uncontrollable for the competing methods when a model is misspecified, a phenomenon routinely encountered in practice.

Dynamic Quantitative Trait Locus Analysis of Plant Phenomic Data

Advanced platforms have recently become available for automatic and systematic quantification of plant growth and development. These new techniques can efficiently produce multiple measurements of phenotypes over time, and introduce time as an extra dimension to quantitative trait locus (QTL) studies. Functional mapping utilizes a class of statistical models for identifying QTLs associated with the growth characteristics of interest. A major benefit of functional mapping is that it integrates information over multiple timepoints, and therefore could increase the statistical power for QTL detection. We review the current development of computationally efficient functional mapping methods which provide invaluable tools for analyzing large-scale timecourse data that are readily available in our post-genome era.

Bayesian hierarchical modelling for inferring genetic interactions in yeast

Quantitative fitness analysis (QFA) is a high throughput experimental and computational methodology for measuring the growth of microbial populations. QFA screens can be used to compare the health of cell populations with and without a mutation in a query gene to infer genetic interaction strengths genomewide, examining thousands of separate genotypes. We introduce Bayesian hierarchical models of population growth rates and genetic interactions that better reflect QFA experimental design than current approaches. Our new approach models population dynamics and genetic interaction simultaneously, thereby avoiding passing information between models via a univariate fitness summary. Matching experimental structure more closely, Bayesian hierarchical approaches use data more efficiently and find new evidence for genes which interact with yeast telomeres within a published data set.

Determination of nonlinear genetic architecture using compressed sensing

Background

One of the fundamental problems of modern genomics is to extract the genetic architecture of a complex trait from a data set of individual genotypes and trait values. Establishing this important connection between genotype and phenotype is complicated by the large number of candidate genes, the potentially large number of causal loci, and the likely presence of some nonlinear interactions between different genes. Compressed Sensing methods obtain solutions to under-constrained systems of linear equations. These methods can be applied to the problem of determining the best model relating genotype to phenotype, and generally deliver better performance than simply regressing the phenotype against each genetic variant, one at a time. We introduce a Compressed Sensing method that can reconstruct nonlinear genetic models (i.e., including epistasis, or gene-gene interactions) from phenotype-genotype (GWAS) data. Our method uses L1-penalized regression applied to nonlinear functions of the sensing matrix.

Results

The computational and data resource requirements for our method are similar to those necessary for reconstruction of linear genetic models (or identification of gene-trait associations), assuming a condition of generalized sparsity, which limits the total number of gene-gene interactions. An example of a sparse nonlinear model is one in which a typical locus interacts with several or even many others, but only a small subset of all possible interactions exist. It seems plausible that most genetic architectures fall in this category. We give theoretical arguments suggesting that the method is nearly optimal in performance, and demonstrate its effectiveness on broad classes of nonlinear genetic models using simulated human genomes and the small amount of currently available real data. A phase transition (i.e., dramatic and qualitative change) in the behavior of the algorithm indicates when sufficient data is available for its successful application.

Conclusion

Our results indicate that predictive models for many complex traits, including a variety of human disease susceptibilities (e.g., with additive heritability h²∼0.5), can be extracted from data sets comprised of n_⋆∼100s individuals, where s is the number of distinct causal variants influencing the trait. For example, given a trait controlled by ∼10 k loci, roughly a million individuals would be sufficient for application of the method.

SIRT6 minor allele genotype is associated with >5-year decrease in lifespan in an aged cohort

Aging is a natural process involving complex interplay between environment, metabolism, and genes. Sirtuin genes and their downstream targets have been associated with lifespan in numerous organisms from nematodes to humans. Several target proteins of the sirtuin genes are key sensors and/or effectors of oxidative stress pathways including FOXO3, SOD3, and AKT1. To examine the relationship between single nucleotide polymorphisms (SNP) at candidate genes in these pathways and human lifespan, we performed a molecular epidemiologic study of an elderly cohort (≥65 years old.). Using age at death as a continuous outcome variable and assuming a co-dominant genetic model within the framework of multi-variable linear regression analysis, the genotype-specific adjusted mean age at death was estimated for individual SNP genotypes while controlling for age-related risk factors including smoking, body mass index, alcohol consumption and co-morbidity. Significant associations were detected between human lifespan and SNPs in genes SIRT3, SIRT5, SIRT6, FOXO3 and SOD3. Individuals with either the CC or CT genotype at rs107251 within SIRT6 displayed >5-year mean survival advantages compared to the TT genotype (5.5 and 5.9 years, respectively; q-value  = 0.012). Other SNPs revealed genotype-specific mean survival advantages ranging from 0.5 to 1.6 years. Gender also modified the effect of SNPs in SIRT3, SIRT5 and AKT1 on lifespan. Our novel findings highlight the impact of sirtuins and sirtuin-related genotypes on lifespan, the importance of evaluating gender and the advantage of using age as a continuous variable in analyses to report mean age at death.

Gene-alcohol interactions identify several novel blood pressure loci including a promising locus near SLC16A9

Alcohol consumption is a known risk factor for hypertension, with recent candidate studies implicating gene-alcohol interactions in blood pressure (BP) regulation. We used 6882 (predominantly) Caucasian participants aged 20-80 years from the Framingham SNP Health Association Resource (SHARe) to perform a genome-wide analysis of SNP-alcohol interactions on BP traits. We used a two-step approach in the ABEL suite to examine genetic interactions with three alcohol measures (ounces of alcohol consumed per week, drinks consumed per week, and the number of days drinking alcohol per week) on four BP traits [systolic (SBP), diastolic (DBP), mean arterial (MAP), and pulse (PP) pressure]. In the first step, we fit a linear mixed model of each BP trait onto age, sex, BMI, and antihypertensive medication while accounting for the phenotypic correlation among relatives. In the second step, we conducted 1 degree-of-freedom (df) score tests of the SNP main effect, alcohol main effect, and SNP-alcohol interaction using the maximum likelihood estimates (MLE) of the parameters from the first step. We then calculated the joint 2 df score test of the SNP main effect and SNP-alcohol interaction using MixABEL. The effect of SNP rs10826334 (near SLC16A9) on SBP was significantly modulated by both the number of alcoholic drinks and the ounces of alcohol consumed per week (p-values of 1.27E-08 and 3.92E-08, respectively). Each copy of the G-allele decreased SBP by 3.79 mmHg in those consuming 14 drinks per week vs. a 0.461 mmHg decrease in non-drinkers. Index SNPs in 20 other loci exhibited suggestive (p-value ≤ 1E-06) associations with BP traits by the 1 df interaction test or joint 2 df test, including 3 rare variants, one low-frequency variant, and SNPs near/in genes ESRRG, FAM179A, CRIPT-SOCS5, KAT2B, ADCY2, GLI3, ZNF716, SLIT1, PDE3A, KERA-LUM, RNF219-AS1, CLEC3A, FBXO15, and IGSF5. SNP-alcohol interactions may enhance discovery of novel variants with large effects that can be targeted with lifestyle modifications.

Exploiting gene expression variation to capture gene-environment interactions for disease

Gene-environment interactions have long been recognized as a fundamental concept in evolutionary, quantitative, and medical genetics. In the genomics era, study of how environment and genome interact to shape gene expression variation is relevant to understanding the genetic architecture of complex phenotypes. While genetic analysis of gene expression variation focused on main effects, little is known about the extent of interaction effects implicating regulatory variants and their consequences on transcriptional variation. Here we survey the current state of the concept of transcriptional gene-environment interactions and discuss its utility for mapping disease phenotypes in light of the insights gained from genome-wide association studies of gene expression.

Hierarchical generalized linear models for multiple groups of rare and common variants: jointly estimating group and individual-variant effects

Complex diseases and traits are likely influenced by many common and rare genetic variants and environmental factors. Detecting disease susceptibility variants is a challenging task, especially when their frequencies are low and/or their effects are small or moderate. We propose here a comprehensive hierarchical generalized linear model framework for simultaneously analyzing multiple groups of rare and common variants and relevant covariates. The proposed hierarchical generalized linear models introduce a group effect and a genetic score (i.e., a linear combination of main-effect predictors for genetic variants) for each group of variants, and jointly they estimate the group effects and the weights of the genetic scores. This framework includes various previous methods as special cases, and it can effectively deal with both risk and protective variants in a group and can simultaneously estimate the cumulative contribution of multiple variants and their relative importance. Our computational strategy is based on extending the standard procedure for fitting generalized linear models in the statistical software R to the proposed hierarchical models, leading to the development of stable and flexible tools. The methods are illustrated with sequence data in gene ANGPTL4 from the Dallas Heart Study. The performance of the proposed procedures is further assessed via simulation studies. The methods are implemented in a freely available R package BhGLM (http://www.ssg.uab.edu/bhglm/).

Methods of plant breeding in the genome era

Methods of genomic value prediction are reviewed. The majority of the methods are related to mixed model methodology, either explicitly or implicitly, by treating systematic environmental effects as fixed and quantitative trait locus (QTL) effects as random. Six different methods are reviewed, including least squares (LS), ridge regression, Bayesian shrinkage, least absolute shrinkage and selection operator (Lasso), empirical Bayes and partial least squares (PLS). The LS and PLS methods are non-Bayesian because they do not require probability distributions for the data. The PLS method is introduced as a special dimension reduction scheme to handle high-density marker information. Theory and methods of cross-validation are described. The leave-one-out cross-validation approach is recommended for model validation. A working example is used to demonstrate the utility of genome selection (GS) in barley. The data set contained 150 double haploid lines and 495 DNA markers covering the entire barley genome, with an average marker interval of 2·23 cM. Eight quantitative traits were included in the analysis. GS using the empirical Bayesian method showed high predictability of the markers for all eight traits with a mean accuracy of prediction of 0·70. With traditional marker-assisted selection (MAS), the average accuracy of prediction was 0·59, giving an average gain of GS over MAS of 0·11. This study provided strong evidence that GS using marker information alone can be an efficient tool for plant breeding.