Improving strategies for detecting genetic patterns of disease susceptibility in association studies

Maternal Age at Menarche Genes Determines Fetal Growth Restriction Risk

We aimed to explore the potential link of maternal age at menarche (mAAM) gene polymorphisms with risk of the fetal growth restriction (FGR). This case (FGR)-control (FGR free) study included 904 women (273 FGR and 631 control) in the third trimester of gestation examined/treated in the Departments of Obstetrics. For single nucleotide polymorphism (SNP) multiplex genotyping, 50 candidate loci of mAAM were chosen. The relationship of mAAM SNPs and FGR was appreciated by regression procedures (logistic/model-based multifactor dimensionality reduction [MB-MDR]) with subsequent in silico assessment of the assumed functionality pithy of FGR-related loci. Three mAAM-appertain loci were FGR-linked to genes such as KISS1 (rs7538038) (effect allele G-odds ratio (OR)allelic = 0.63/pperm = 0.0003; ORadditive = 0.61/pperm = 0.001; ORdominant = 0.56/pperm = 0.001), NKX2-1 (rs999460) (effect allele A-ORallelic = 1.37/pperm = 0.003; ORadditive = 1.45/pperm = 0.002; ORrecessive = 2.41/pperm = 0.0002), GPRC5B (rs12444979) (effect allele T-ORallelic = 1.67/pperm = 0.0003; ORdominant = 1.59/pperm = 0.011; ORadditive = 1.56/pperm = 0.009). The haplotype ACA FSHB gene (rs555621*rs11031010*rs1782507) was FRG-correlated (OR = 0.71/pperm = 0.05). Ten FGR-implicated interworking models were founded for 13 SNPs (pperm ≤ 0.001). The rs999460 NKX2-1 and rs12444979 GPRC5B interplays significantly influenced the FGR risk (these SNPs were present in 50% of models). FGR-related mAAM-appertain 15 polymorphic variants and 350 linked SNPs were functionally momentous in relation to 39 genes participating in the regulation of hormone levels, the ovulation cycle process, male gonad development and vitamin D metabolism. Thus, this study showed, for the first time, that the mAAM-appertain genes determine FGR risk.

The Impact of Genetic Polymorphisms in Glutamate-Cysteine Ligase, a Key Enzyme of Glutathione Biosynthesis, on Ischemic Stroke Risk and Brain Infarct Size

The purpose of this pilot study was to explore whether polymorphisms in genes encoding the catalytic (GCLC) and modifier (GCLM) subunits of glutamate-cysteine ligase, a rate-limiting enzyme in glutathione synthesis, play a role in the development of ischemic stroke (IS) and the extent of brain damage. A total of 1288 unrelated Russians, including 600 IS patients and 688 age- and sex-matched healthy subjects, were enrolled for the study. Nine common single nucleotide polymorphisms (SNPs) of the GCLC and GCLM genes were genotyped using the MassArray-4 system. SNP rs2301022 of GCLM was strongly associated with a decreased risk of ischemic stroke regardless of sex and age (OR = 0.39, 95%CI 0.24−0.62, p < 0.0001). Two common haplotypes of GCLM possessed protective effects against ischemic stroke risk (p < 0.01), but exclusively in nonsmoker patients. Infarct size was increased by polymorphisms rs636933 and rs761142 of GCLC. The mbmdr method enabled identifying epistatic interactions of GCLC and GCLM gene polymorphisms with known IS susceptibility genes that, along with environmental risk factors, jointly contribute to the disease risk and brain infarct size. Understanding the impact of genes and environmental factors on glutathione metabolism will allow the development of effective strategies for the treatment of ischemic stroke and disease prevention.

Detecting gene-gene interactions from GWAS using diffusion kernel principal components

Genes and gene products do not function in isolation but as components of complex networks of macromolecules through physical or biochemical interactions. Dependencies of gene mutations on genetic background (i.e., epistasis) are believed to play a role in understanding molecular underpinnings of complex diseases such as inflammatory bowel disease (IBD). However, the process of identifying such interactions is complex due to for instance the curse of high dimensionality, dependencies in the data and non-linearity. Here, we propose a novel approach for robust and computationally efficient epistasis detection. We do so by first reducing dimensionality, per gene via diffusion kernel principal components (kpc). Subsequently, kpc gene summaries are used for downstream analysis including the construction of a gene-based epistasis network. We show that our approach is not only able to recover known IBD associated genes but also additional genes of interest linked to this difficult gastrointestinal disease.

Comprehensive Statistical and Bioinformatics Analysis in the Deciphering of Putative Mechanisms by Which Lipid-Associated GWAS Loci Contribute to Coronary Artery Disease

The study was designed to evaluate putative mechanisms by which lipid-associated loci identified by genome-wide association studies (GWAS) are involved in the molecular pathogenesis of coronary artery disease (CAD) using a comprehensive statistical and bioinformatics analysis. A total of 1700 unrelated individuals of Slavic origin from the Central Russia, including 991 CAD patients and 709 healthy controls were examined. Sixteen lipid-associated GWAS loci were selected from European studies and genotyped using the MassArray-4 system. The polymorphisms were associated with plasma lipids such as total cholesterol (rs12328675, rs4846914, rs55730499, and rs838880), LDL-cholesterol (rs3764261, rs55730499, rs1689800, and rs838880), HDL-cholesterol (rs3764261) as well as carotid intima-media thickness/CIMT (rs12328675, rs11220463, and rs1689800). Polymorphisms such as rs4420638 of APOC1 (p = 0.009), rs55730499 of LPA (p = 0.0007), rs3136441 of F2 (p < 0.0001), and rs6065906 of PLTP (p = 0.002) showed significant associations with the risk of CAD, regardless of sex, age, and body mass index. A majority of the observed associations were successfully replicated in large independent cohorts. Bioinformatics analysis allowed establishing (1) phenotype-specific and shared epistatic gene–gene and gene–smoking interactions contributing to all studied cardiovascular phenotypes; (2) lipid-associated GWAS loci might be allele-specific binding sites for transcription factors from gene regulatory networks controlling multifaceted molecular mechanisms of atherosclerosis.

Polymorphisms of the TNF, LTA, and TNFRSF1B genes are associated with onsets of menarche and menopause in US women of European ancestry

BACKGROUND

The TNF, LTA and TNFRSF1B genes have been implicated in various traits related to menarche and menopause.

AIM

To analyse the TNF, LTA and TNFRSF1B genes for their association with ages at menarche (AM) and natural menopause (ANM).

SUBJECTS AND METHODS

The study sample consisted of 314 unrelated females of European ancestry. Twenty SNPs located in and near the genes were analysed using various statistical methods. In addition, the functional significance of the loci associated with AM and ANM was analysed in silico.

RESULTS

Locus rs2229094 of the LTA gene was associated with AM according to the additive (β = -0.295, p_perm = 0.016) and recessive (β = -0.940, p_perm = 0.016) genetic models. Haplotype GG rs1148459-rs590368 of the TNFRSF1B gene was associated with AM (β = 0.307, p_perm = 0.023). Haplotype GCA rs2844484-rs2229094-rs1799964 was associated with ANM after adjustment for covariates (β = -1.020, p_perm = 0.035). All studied loci were associated with ANM after adjustment for breastfeeding (raw p < 0.05). In addition, eight of the most significant models of interlocus interactions were associated with AM and five with ANM.

CONCLUSION

The results of the present study suggest that the TNF, LTA and TNFRSF1B genes are associated with AM and ANM.

The Link between Type 2 Diabetes Mellitus and the Polymorphisms of Glutathione-Metabolizing Genes Suggests a New Hypothesis Explaining Disease Initiation and Progression

The present study investigated whether type 2 diabetes (T2D) is associated with polymorphisms of genes encoding glutathione-metabolizing enzymes such as glutathione synthetase (GSS) and gamma-glutamyl transferase 7 (GGT7). A total of 3198 unrelated Russian subjects including 1572 T2D patients and 1626 healthy subjects were enrolled. Single nucleotide polymorphisms (SNPs) of the GSS and GGT7 genes were genotyped using the MassArray-4 system. We found that the GSS and GGT7 gene polymorphisms alone and in combinations are associated with T2D risk regardless of sex, age, and body mass index, as well as correlated with plasma glutathione, hydrogen peroxide, and fasting blood glucose levels. Polymorphisms of GSS (rs13041792) and GGT7 (rs6119534 and rs11546155) genes were associated with the tissue-specific expression of genes involved in unfolded protein response and the regulation of proteostasis. Transcriptome-wide association analysis has shown that the pancreatic expression of some of these genes such as EDEM2, MYH7B, MAP1LC3A, and CPNE1 is linked to the genetic risk of T2D. A comprehensive analysis of the data allowed proposing a new hypothesis for the etiology of type 2 diabetes that endogenous glutathione deficiency might be a key condition responsible for the impaired folding of proinsulin which triggered an unfolded protein response, ultimately leading to beta-cell apoptosis and disease development.

Germline Variants and Genetic Interactions of Several EMT Regulatory Genes Increase the Risk of HBV-Related Hepatocellular Carcinoma

Epithelial-mesenchymal transition (EMT) plays an important role in the development of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC). We hypothesized that germline variants in the major EMT regulatory genes (SNAIL1, ZEB1, ZEB2, TWIST1) may influence the development of HBV-related HCC. We included 421 cases of HBsAg-positive patients with HCC, 1371 cases of HBsAg-positive subjects without HCC [patients with chronic hepatitis B (CHB) or liver cirrhosis (LC)] and 618 cases of healthy controls in the case-control study. Genotype, allele, and haplotype associations in the major EMT regulatory genes were tested. Environment-gene and gene-gene interactions were analysed using the non-parametric model-free multifactor dimensionality reduction (MDR) method. The SNAIL1rs4647958T>C was associated with a significantly increased risk of both HCC (CT+CC vs. TT: OR=1.559; 95% confidence interval [CI], 1.073-2.264; P=0.020) and CHB+LC (CT+CC vs. TT: OR=1.509; 95% CI, 1.145-1.988; P=0.003). Carriers of the TWIST1rs2285681G>C (genotypes CT+CC) had an increased risk of HCC (CG+CC vs. GG: OR=1.407; 95% CI, 1.065-1.858; P=0.016). The ZEB2rs3806475T>C was associated with significantly increased risk of both HCC (P _recessive =0.001) and CHB+LC (P _recessive<0.001). The CG haplotype of the rs4647958/rs1543442 haplotype block was associated with significant differences between healthy subjects and HCC patients (P=0.0347). Meanwhile, the CT haplotype of the rs2285681/rs2285682 haplotype block was associated with significant differences between CHB+LC and HCC patients (P=0.0123). In MDR analysis, the combination of TWIST1rs2285681, ZEB2rs3806475, SNAIL1rs4647958 exhibited the most significant association with CHB+LC and Health control in the three-locus model. Our results suggest significant single-gene associations and environment-gene/gene-gene interactions of EMT-related genes with HBV-related HCC.

Polymorphisms of the matrix metalloproteinase genes are associated with essential hypertension in a Caucasian population of Central Russia

This study aimed to determine possible association of eight polymorphisms of seven MMP genes with essential hypertension (EH) in a Caucasian population of Central Russia. Eight SNPs of the MMP1, MMP2, MMP3, MMP7, MMP8, MMP9, and MMP12 genes and their gene–gene (epistatic) interactions were analyzed for association with EH in a cohort of 939 patients and 466 controls using logistic regression and assuming additive, recessive, and dominant genetic models. The functional significance of the polymorphisms associated with EH and 114 variants linked to them (r² ≥ 0.8) was analyzed in silico. Allele G of rs11568818 MMP7 was associated with EH according to all three genetic models (OR = 0.58–0.70, p_perm = 0.01–0.03). The above eight SNPs were associated with the disorder within 12 most significant epistatic models (OR = 1.49–1.93, p_perm < 0.02). Loci rs1320632 MMP8 and rs11568818 MMP7 contributed to the largest number of the models (12 and 10, respectively). The EH-associated loci and 114 SNPs linked to them had non-synonymous, regulatory, and eQTL significance for 15 genes, which contributed to the pathways related to metalloendopeptidase activity, collagen degradation, and extracellular matrix disassembly. In summary, eight studied SNPs of MMPs genes were associated with EH in the Caucasian population of Central Russia.

Empowering individual trait prediction using interactions for precision medicine

Background

One component of precision medicine is to construct prediction models with their predicitve ability as high as possible, e.g. to enable individual risk prediction. In genetic epidemiology, complex diseases like coronary artery disease, rheumatoid arthritis, and type 2 diabetes, have a polygenic basis and a common assumption is that biological and genetic features affect the outcome under consideration via interactions. In the case of omics data, the use of standard approaches such as generalized linear models may be suboptimal and machine learning methods are appealing to make individual predictions. However, most of these algorithms focus mostly on main or marginal effects of the single features in a dataset. On the other hand, the detection of interacting features is an active area of research in the realm of genetic epidemiology. One big class of algorithms to detect interacting features is based on the multifactor dimensionality reduction (MDR). Here, we further develop the model-based MDR (MB-MDR), a powerful extension of the original MDR algorithm, to enable interaction empowered individual prediction.

Results

Using a comprehensive simulation study we show that our new algorithm (median AUC: 0.66) can use information hidden in interactions and outperforms two other state-of-the-art algorithms, namely the Random Forest (median AUC: 0.54) and Elastic Net (median AUC: 0.50), if interactions are present in a scenario of two pairs of two features having small effects. The performance of these algorithms is comparable if no interactions are present. Further, we show that our new algorithm is applicable to real data by comparing the performance of the three algorithms on a dataset of rheumatoid arthritis cases and healthy controls. As our new algorithm is not only applicable to biological/genetic data but to all datasets with discrete features, it may have practical implications in other research fields where interactions between features have to be considered as well, and we made our method available as an R package (https://github.com/imbs-hl/MBMDRClassifieR).

Conclusions

The explicit use of interactions between features can improve the prediction performance and thus should be included in further attempts to move precision medicine forward.

Candidate Genes for Age at Menarche Are Associated With Uterine Leiomyoma

Age at menarche (AAM) is an important marker of the pubertal development and function of the hypothalamic-pituitary-ovarian system. It was reported as a possible factor for a risk of uterine leiomyoma (UL). However, while more than 350 loci for AAM have been determined by genome-wide association studies (GWASs) to date, no studies of these loci for their association with UL have been conducted so far. In this study, we analyzed 52 candidate loci for AAM for possible association with UL in a sample of 569 patients and 981 controls. The results of the study suggested that 23 out of the 52 studied polymorphisms had association with UL. Locus rs7759938 LIN28B was individually associated with the disease according to the dominant model. Twenty loci were associated with UL within 11 most significant models of intergenic interactions. Nine loci involved in 16 most significant models of interactions between single-nucleotide polymorphism (SNP), induced abortions, and chronic endometritis were associated with UL. Among the 23 loci associated with UL, 16 manifested association also with either AAM (7 SNPs) or height and/or body mass index (BMI) (13 SNPs). The above 23 SNPs and 514 SNPs linked to them have non-synonymous, regulatory, and expression quantitative trait locus (eQTL) significance for 35 genes, which play roles in the pathways related to development of the female reproductive organs and hormone-mediated signaling [false discovery rate (FDR) ≤ 0.05]. This is the first study reporting associations of candidate genes for AAM with UL.

Machine learning in prediction of genetic risk of nonsyndromic oral clefts in the Brazilian population

OBJECTIVES

Genetic variants in multiple genes and loci have been associated with the risk of nonsyndromic cleft lip with or without cleft palate (NSCL ± P). However, the estimation of risk remains challenge, because most of these variants are population-specific rendering the identification of the underlying genetic risk difficult. Herein we examined the use of machine learning network in previously reported single nucleotide polymorphisms (SNPs) to predict risk of NSCL ± P in the Brazilian population.

MATERIALS AND METHODS

Random forest and neural network methods were applied in 72 SNPs in a case-control sample composed by 722 NSCL ± P and 866 controls for discrimination of NSCL ± P risk. SNP-SNP interactions and functional annotation biological processes associated with the identified NSCL ± P risk genes were verified.

RESULTS

Supervised random forest decision trees revealed high scores of importance for the SNPs rs11717284 and rs1875735 in FGF12, rs41268753 in GRHL3, rs2236225 in MTHFD1, rs2274976 in MTHFR, rs2235371 and rs642961 in IRF6, rs17085106 in RHPN2, rs28372960 in TCOF1, rs7078160 in VAX1, rs10762573 and rs2131960 in VCL, and rs227731 in 17q22, with an accuracy of 99% and an error rate of approximately 3% to predict the risk of NSCL ± P. Those same 13 SNPs were considered the most important for the neural network to effectively predict NSCL ± P risk, with an overall accuracy of 94%. Multivariate regression model revealed significant interactions among all SNPs, with an exception of those in FGF12 and MTHFD1. The most significantly biological processes for selected genes were those involved in tissue and epithelium development; neural tube closure; and metabolism of methionine, folate, and homocysteine.

CONCLUSIONS

Our results provide novel clues for genetic mechanism studies of NSCL ± P and point out for a machine learning model composed by 13 SNPs that is capable of predicting NSCL ± P risk.

CLINICAL RELEVANCE

Although validation is necessary, this genetic panel can be useful in the near future to assist in NSCL ± P genetic counseling.

Combination of Ensembles of Regularized Regression Models with Resampling-Based Lasso Feature Selection in High Dimensional Data

In high-dimensional data, the performances of various classifiers are largely dependent on the selection of important features. Most of the individual classifiers with the existing feature selection (FS) methods do not perform well for highly correlated data. Obtaining important features using the FS method and selecting the best performing classifier is a challenging task in high throughput data. In this article, we propose a combination of resampling-based least absolute shrinkage and selection operator (LASSO) feature selection (RLFS) and ensembles of regularized regression (ERRM) capable of dealing data with the high correlation structures. The ERRM boosts the prediction accuracy with the top-ranked features obtained from RLFS. The RLFS utilizes the lasso penalty with sure independence screening (SIS) condition to select the top k ranked features. The ERRM includes five individual penalty based classifiers: LASSO, adaptive LASSO (ALASSO), elastic net (ENET), smoothly clipped absolute deviations (SCAD), and minimax concave penalty (MCP). It was built on the idea of bagging and rank aggregation. Upon performing simulation studies and applying to smokers’ cancer gene expression data, we demonstrated that the proposed combination of ERRM with RLFS achieved superior performance of accuracy and geometric mean.

Significant role of gene-gene interactions of clock genes in mood disorder

BACKGROUND

The genetic interactions in the circadian rhythm biological system are promising as a source of pathophysiology in mood disorder. We examined the role of the gene-gene interactions of clock genes in mood disorder.

METHODS

We included 413 patients with mood disorder and 1294 controls. The clock genes investigated were BHLHB2, CLOCK, CSNK1E, NR1D1, PER2, PER3, and TIMELESS. Allele, genotype, and haplotype associations were tested. Gene--gene interactions were analyzed using the non-parametric model-free multifactor-dimensionality reduction (MDR) method.

RESULTS

TIMELESS rs4630333 and CSNK1E rs135745 were significantly associated with both major depressive disorder and bipolar disorder. The CLOCK haplotype was also strongly associated. The genetic roles of these SNPs were consistent from the allele and genotypic associations to the MDR interaction results. In MDR analysis, the combination of TIMELESS rs4630333 and CSNK1E rs135745 exhibited the most significant association with mood disorders in the two-locus model. BHLHB2 rs2137947 for major depressive disorder and CLOCK rs12649507 for bipolar disorder were the most significant third loci in the three-locus combination model. The four-locus SNP combination model showed the best balanced accuracy (BA), but its cross-validation consistency (CVC) was unsatisfactory.

LIMITATIONS

We included only 17 SNPs for seven circadian genes due to our limited resources; all subjects were ethnically Korean.

CONCLUSIONS

Our results suggest significant single-gene associations and gene-gene interactions of circadian genes with mood disorder. Gene-gene interactions play a crucial role in mood disorder, even when individual clock genes do not have significant roles.

Confounding of linkage disequilibrium patterns in large scale DNA based gene-gene interaction studies

Background

In Genome-Wide Association Studies (GWAS), the concept of linkage disequilibrium is important as it allows identifying genetic markers that tag the actual causal variants. In Genome-Wide Association Interaction Studies (GWAIS), similar principles hold for pairs of causal variants. However, Linkage Disequilibrium (LD) may also interfere with the detection of genuine epistasis signals in that there may be complete confounding between Gametic Phase Disequilibrium (GPD) and interaction. GPD may involve unlinked genetic markers, even residing on different chromosomes. Often GPD is eliminated in GWAIS, via feature selection schemes or so-called pruning algorithms, to obtain unconfounded epistasis results. However, little is known about the optimal degree of GPD/LD-pruning that gives a balance between false positive control and sufficient power of epistasis detection statistics. Here, we focus on Model-Based Multifactor Dimensionality Reduction as one large-scale epistasis detection tool. Its performance has been thoroughly investigated in terms of false positive control and power, under a variety of scenarios involving different trait types and study designs, as well as error-free and noisy data, but never with respect to multicollinear SNPs.

Results

Using real-life human LD patterns from a homogeneous subpopulation of British ancestry, we investigated the impact of LD-pruning on the statistical sensitivity of MB-MDR. We considered three different non-fully penetrant epistasis models with varying effect sizes. There is a clear advantage in pre-analysis pruning using sliding windows at r² of 0.75 or lower, but using a threshold of 0.20 has a detrimental effect on the power to detect a functional interactive SNP pair (power < 25%). Signal sensitivity, directly using LD-block information to determine whether an epistasis signal is present or not, benefits from LD-pruning as well (average power across scenarios: 87%), but is largely hampered by functional loci residing at the boundaries of an LD-block.

Conclusions

Our results confirm that LD patterns and the position of causal variants in LD blocks do have an impact on epistasis detection, and that pruning strategies and LD-blocks definitions combined need careful attention, if we wish to maximize the power of large-scale epistasis screenings.

Two-Stage Classification with SIS Using a New Filter Ranking Method in High Throughput Data

Over the last decade, high dimensional data have been popularly paid attention to in bioinformatics. These data increase the likelihood of detecting the most promising novel information. However, there are limitations of high-performance computing and overfitting issues. To overcome the issues, alternative strategies need to be explored for the detection of true important features. A two-stage approach, filtering and variable selection steps, has been receiving attention. Filtering methods are divided into two categories of individual ranking and feature subset selection methods. Both have issues with the lack of consideration for joint correlation among features and computing time of an NP-hard problem. Therefore, we proposed a new filter ranking method (PF) using the elastic net penalty with sure independence screening (SIS) based on resampling technique to overcome these issues. We demonstrated that SIS-LASSO, SIS-MCP, and SIS-SCAD with the proposed filtering method achieved superior performance of not only accuracy, AUROC, and geometric mean but also true positive detection compared to those with the marginal maximum likelihood ranking method (MMLR) through extensive simulation studies. In addition, we applied it in a real application of colon and lung cancer gene expression data to investigate the classification performance and power of detecting true genes associated with colon and lung cancer.

Genetic associations of the vitamin D and antiviral pathways with natural resistance to HIV-1 infection are influenced by interpopulation variability

Vitamin D (VitD) may modulate anti-HIV-1 responses modifying the risk to acquire the HIV-1-infection. We performed a nested case-control exploratory study involving 413 individuals; HIV-1-exposed seropositives (cases) and seronegatives (HESN) (controls) from three cohorts: sexually-exposed from Colombia and Italy and parenterally-exposed from Spain. The association and interactions of 139 variants in 9 VitD pathway genes, and in 14 antiviral genes with resistance/susceptibility (R/S) to HIV-1 infection was evaluated. Associations between variants and mRNA levels were also analyzed in the Colombian samples. Variants and haplotypes in genes of VitD and antiviral pathways were associated with R/S, but specific associations were not reproduced in all cohorts. Allelic heterogeneity could explain such inconsistency since the associations found in all cohorts were consistently in the same genes: VDR and RXRA of the VitD pathway genes and in TLR2 and RNASE4. Remarkably, the multi-locus genotypes (interacting variants) observed in genes of VitD and antiviral pathways were present in most HESNs of all cohorts. Finally, HESNs carrying resistance-associated variants had higher levels of VitD in plasma, of VDR mRNA in blood cells, and of ELAFIN and defensins mRNA in the oral mucosa. In conclusion, despite allelic heterogeneity, most likely due to differences in the genetic history of the populations, the associations were locus dependent suggesting that genes of the VitD pathway might act in concert with antiviral genes modulating the resistance phenotype of the HESNs. Although these associations were significant after permutation test, only haplotype results remained statistically significant after Bonferroni test, requiring further replications in larger cohorts and functional analyzes to validate these conclusions.

Matrix metalloproteinases as target genes for gene regulatory networks driving molecular and cellular pathways related to a multistep pathogenesis of cerebrovascular disease

The present study investigated a joint contribution of matrix metalloproteinases (MMPs) genes to ischemic stroke (IS) development and analyzed interactions between MMP genes and genome-wide associated loci for IS. A total of 1288 unrelated Russians (600 IS patients and 688 healthy individuals) from Central Russia were recruited for the study. Genotyping of seven single nucleotide polymorphisms (SNPs) of MMP genes (rs1799750, rs243865, rs3025058, rs11225395, rs17576, rs486055, and rs2276109) and eight genome-wide associated loci for IS were done using Taq-Man-based assays and MALDI-TOF mass spectrometry iPLEX platform, respectively. Allele - 799T at rs11225395 of the MMP8 gene was significantly associated with a decreased risk of IS after adjustment for sex and age (OR = 0.82; 95%CI, 0.70-0.96; P = 0.016). The model-based multifactor dimensionality reduction method has revealed 21 two-order, 124 three-order, and 474 four-order gene-gene (G×G) interactions models meaningfully (P_perm  < 0.05) associated with the IS risk. The bioinformatic analysis enabled establishing the studied MMP gene polymorphisms possess a clear regulatory potential and may be targeted by gene regulatory networks driving molecular and cellular pathways related to the pathogenesis of IS. In conclusion, the present study was the first to identify an association between polymorphism rs11225395 of the MMP8 gene and IS risk. The study findings also indicate that MMPs deserve special attention as a potential class of genes influencing the multistep mechanisms of cerebrovascular disease including atherosclerosis in cerebral arteries, acute cerebral artery occlusion as well as the ischemic injury of the brain and its recovery.

Human collectin-11 (COLEC11) and its synergic genetic interaction with MASP2 are associated with the pathophysiology of Chagas Disease

Chagas Disease (CD) is an anthropozoonosis caused by Trypanosoma cruzi. With complex pathophysiology and variable clinical presentation, CD outcome can be influenced by parasite persistence and the host immune response. Complement activation is one of the primary defense mechanisms against pathogens, which can be initiated via pathogen recognition by pattern recognition molecules (PRMs). Collectin-11 is a multifunctional soluble PRM lectin, widely distributed throughout the body, with important participation in host defense, homeostasis, and embryogenesis. In complex with mannose-binding lectin-associated serine proteases (MASPs), collectin-11 may initiate the activation of complement, playing a role against pathogens, including T. cruzi. In this study, collectin-11 plasma levels and COLEC11 variants in exon 7 were assessed in a Brazilian cohort of 251 patients with chronic CD and 108 healthy controls. Gene-gene interactions between COLEC11 and MASP2 variants were analyzed. Collectin-11 levels were significantly decreased in CD patients compared to controls (p<0.0001). The allele rs7567833G, the genotypes rs7567833AG and rs7567833GG, and the COLEC11*GGC haplotype were related to T. cruzi infection and clinical progression towards symptomatic CD. COLEC11 and MASP2*CD risk genotypes were associated with cardiomyopathy (p = 0.014; OR 9.3, 95% CI 1.2–74) and with the cardiodigestive form of CD (p = 0.005; OR 15.2, 95% CI 1.7–137), suggesting that both loci act synergistically in immune modulation of the disease. The decreased levels of collectin-11 in CD patients may be associated with the disease process. The COLEC11 variant rs7567833G and also the COLEC11 and MASP2*CD risk genotype interaction were associated with the pathophysiology of CD.

The heterogeneity of clinical progression during chronic Trypanosoma cruzi infection and the mechanisms determining why some individuals develop symptoms whereas others remain asymptomatic are still poorly understood. The pathogenesis of chronic Chagas Disease (CD) has been attributed mainly to the persistence of the causing parasite and the character of individual host immune responses. Collectin-11 is a host immune response molecule with affinity for sugars found on the T. cruzi’s surface. Together with mannose-binding lectin-associated serine proteases (MASPs), it triggers the host defense response against pathogens. Genetic variants and protein levels of MASP-2 and the mannose-binding lectin (MBL), a molecule structurally similar to collectin-11, have been found to be associated with susceptibility to T. cruzi infection and clinical progression to cardiomyopathy. This prompted us to investigate collectin-11 genetic variants and protein levels in 251 patients with chronic CD and 108 healthy individuals, and to examine the effect of gene interaction between COLEC11 and MASP2 risk mutations. We found an association to CD infection with COLEC11 gene variants and reduced collectin-11 levels. The concomitant presence of these genetic variants and MASP2 risk mutations greatly increased the odds for cardiomyopathy. This is the first study to reveal a role for collectin-11 and COLEC11-MASP2 gene interaction in the pathogenesis of CD.

Gene-Gene and Gene-Environment Interactions

Identifying gene-gene and gene-environment interactions may help us to better describe the genetic architecture for complex traits. While advances have been made in identifying genetic variants associated with complex traits through more dense panels of genetic variants and larger sample sizes, genome-wide interaction analyses are still limited in power to detect interactions with small effect sizes, rare frequencies, and higher order interactions. This chapter outlines methods for detecting both gene-gene and gene-environment interactions both through explicit tests for interactions (i.e., ones in which the interaction is tested directly) and non-explicit tests (i.e., ones in which an interaction is allowed for in the test, but does not test for the interaction directly) as well as approaches for increasing power by reducing the search space. Issues relating to multiple test correction, replication, and the reporting of interaction results in publications.

Relationships of SLC2A4, RBP4, PCK1, and PI3K Gene Polymorphisms with Gestational Diabetes Mellitus in a Chinese Population

Background

Solute carrier family 2 member 4- (SLC2A4-) retinol binding protein-4- (RBP4-) phosphoenolpyruvate carboxykinase 1 (PCK1)/phosphoinositide 3-kinase (PI3K) is an adipocyte derived “signalling pathway” that may contribute to the pathogenesis of type 2 diabetes mellitus (T2DM). We explored whether single nucleotide polymorphisms (SNPs) of these “signalling pathway” genes are associated with gestational diabetes mellitus (GDM).

Methods

Case-control studies were conducted to compare GDM and control groups. A total of 334 cases and 367 controls were recruited. Seventeen candidate SNPs of the pathway were selected. Chi-square tests, logistic regression, and linear regression were used to estimate the relationships of SNPs with GDM risk and oral glucose tolerance test (OGTT), fasting insulin, and homeostasis model assessment of insulin resistance (HOMA-IR) levels. Model-based multifactor dimensionality reduction was used to estimate the adjusted interactions between genes. Regression and interaction analyses were adjusted by maternal age, prepregnancy BMI, and weekly BMI growth. The Bonferroni correction was applied for multiple comparisons.

Results

RBP4 rs7091052 was significantly associated with GDM risk. SLC2A4 rs5435, RBP4 rs7091052, PCK1 rs1042531 and rs2236745, and PIK3R1 (coding gene of the PI3K P85 subunit) rs34309 were associated with OGTT, fasting insulin, and HOMA-IR levels in the linear regression analysis. The gene-gene interaction analysis showed that, compared with pregnant women with other genotype combinations, women with SLC2A4 rs5435 (CC/CT), RBP4 rs7091052 (CC), PCK1 rs1042531 (TT/TG) and rs2236745 (TT), and PIK3R1 rs34309 (AA) had lower GDM risk.

Conclusion

SLC2A4, RBP4, PCK1, and PIK3R1 genes may be involved in the pathogenesis of GDM.

Association of genetic polymorphisms with age at menarche in Russian women

OBJECTIVES

Examine the association of genetic polymorphisms with age at menarche (AAM) in Russian women.

STUDY DESIGN

A total of 1613 Russian females were recruited for the study. Fifty two polymorphisms were analyzed for their association with AAM, height, and BMI. The associations were analyzed assuming the additive, dominant, and recessive models and using the log-linear regression as implemented in PLINK v. 2.050. The 2-, 3-, and 4-loci models of gene-gene interactions were analyzed using the MB-MDR method and validated by the permutation test.

MAIN OUTCOME MEASURES

Genetic polymorphism rs6438424 3q13.32 was independently associated with AAM in Russian women. In addition, 14 SNPs were determined as possible contributors to this trait through gene-gene interactions.

RESULTS

The obtained results suggest that 14 out of 52 studied polymorphisms may contribute to AAM in Russian women. The rs6438424 3q13.32 polymorphism was associated with AAM according to both additive and dominant models (р_perm = 0.005). In total 12 two-, three-, and four-locus models of gene-gene interactions were determined as contributing to AAM (p_perm ≤ 0.006). Nine of the 14 AAM-associated SNPs are also associated with height and BMI (p_perm ≤ 0.003). Among 14 AAM-associated SNPs (a priori all having regulatory significance), the highest regulatory potential was determined for rs4633 COMT, rs2164808 POMC, rs2252673INSR, rs6438424 3q13.32, and rs10769908 STK33. Eleven loci are cis-eQTL and affect expression of 14 genes in various tissues and organs (FDR < 0.05). The neuropeptide-encoding genes were overrepresented among the AAM-associated genes (p_bonf = 0.039).

CONCLUSIONS

The rs6438424 polymorphism is independently associated with AAM in Russian females in this study. The other 14 SNPs manifest this association through gene-gene interactions.

HisCoM-GGI: Hierarchical structural component analysis of gene-gene interactions

Although genome-wide association studies (GWAS) have successfully identified thousands of single nucleotide polymorphisms (SNPs) associated with common diseases, these observations are limited for fully explaining "missing heritability". Determining gene-gene interactions (GGI) are one possible avenue for addressing the missing heritability problem. While many statistical approaches have been proposed to detect GGI, most of these focus primarily on SNP-to-SNP interactions. While there are many advantages of gene-based GGI analyses, such as reducing the burden of multiple-testing correction, and increasing power by aggregating multiple causal signals across SNPs in specific genes, only a few methods are available. In this study, we proposed a new statistical approach for gene-based GGI analysis, "Hierarchical structural CoMponent analysis of Gene-Gene Interactions" (HisCoM-GGI). HisCoM-GGI is based on generalized structured component analysis, and can consider hierarchical structural relationships between genes and SNPs. For a pair of genes, HisCoM-GGI first effectively summarizes all possible pairwise SNP-SNP interactions into a latent variable, from which it then performs GGI analysis. HisCoM-GGI can evaluate both gene-level and SNP-level interactions. Through simulation studies, HisCoM-GGI demonstrated higher statistical power than existing gene-based GGI methods, in analyzing a GWAS of a Korean population for identifying GGI associated with body mass index. Resultantly, HisCoM-GGI successfully identified 14 potential GGI, two of which, (NCOR2 <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mo>×</mml:mo></mml:math> SPOCK1) and (LINGO2 <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mo>×</mml:mo></mml:math> ZNF385D) were successfully replicated in independent datasets. We conclude that HisCoM-GGI method may be a valuable tool for genome to identify GGI in missing heritability, allowing us to better understand the biological genetic mechanisms of complex traits. We conclude that HisCoM-GGI method may be a valuable tool for genome to identify GGI in missing heritability, allowing us to better understand biological genetic mechanisms of complex traits. An implementation of HisCoM-GGI can be downloaded from the website ( http://statgen.snu.ac.kr/software/hiscom-ggi ).

Transferring entropy to the realm of GxG interactions

Genome-wide association studies are moving to genome-wide interaction studies, as the genetic background of many diseases appears to be more complex than previously supposed. Thus, many statistical approaches have been proposed to detect gene-gene (GxG) interactions, among them numerous information theory-based methods, inspired by the concept of entropy. These are suggested as particularly powerful and, because of their nonlinearity, as better able to capture nonlinear relationships between genetic variants and/or variables. However, the introduced entropy-based estimators differ to a surprising extent in their construction and even with respect to the basic definition of interactions. Also, not every entropy-based measure for interaction is accompanied by a proper statistical test. To shed light on this, a systematic review of the literature is presented answering the following questions: (1) How are GxG interactions defined within the framework of information theory? (2) Which entropy-based test statistics are available? (3) Which underlying distribution do the test statistics follow? (4) What are the given strengths and limitations of these test statistics?

Epistatic interaction of apolipoprotein E and lipolysis-stimulated lipoprotein receptor genetic variants is associated with Alzheimer's disease

The ε4 allele of the apolipoprotein E (APOE) gene common polymorphism is the strongest genetic risk factor for Alzheimer's disease (AD). Human APOE gene is located on chromosome 19q13.1, a region linked to AD that also includes the LSR gene, which encodes the lipolysis-stimulated lipoprotein receptor (LSR). As an APOE receptor, LSR is involved in the regulation of lipid homeostasis in both periphery and brain. This study aimed to determine the potential interactions between 2 LSR genetic variants, rs34259399 and rs916147, and the APOE common polymorphism in 142 AD subjects (mean age: 73.16 ± 8.50 years) and 63 controls (mean age: 70.41 ± 8.49 years). A significant epistatic interaction was observed between APOE and both LSR variants, rs34259399 (beta = -0.95; p = 2 × 10^-5) and rs916147 (beta = -0.83; p = 6.8 × 10^-3). Interestingly, the interaction of LSR polymorphisms with APOE non-ε4 alleles increased AD risk. This indicates the existence of complex molecular interactions between these 2 neighboring genes involved in the pathogenesis of AD, which merits further investigation.

A Review on Methods for Detecting SNP Interactions in High-Dimensional Genomic Data

In this era of genome-wide association studies (GWAS), the quest for understanding the genetic architecture of complex diseases is rapidly increasing more than ever before. The development of high throughput genotyping and next generation sequencing technologies enables genetic epidemiological analysis of large scale data. These advances have led to the identification of a number of single nucleotide polymorphisms (SNPs) responsible for disease susceptibility. The interactions between SNPs associated with complex diseases are increasingly being explored in the current literature. These interaction studies are mathematically challenging and computationally complex. These challenges have been addressed by a number of data mining and machine learning approaches. This paper reviews the current methods and the related software packages to detect the SNP interactions that contribute to diseases. The issues that need to be considered when developing these models are addressed in this review. The paper also reviews the achievements in data simulation to evaluate the performance of these models. Further, it discusses the future of SNP interaction analysis.

Interactions Between PPARG and AGTR1 Gene Polymorphisms on the Risk of Hypertension in Chinese Han Population

AIMS

To explore the interactions between PPARG and AGTR1 polymorphisms and their associations with hypertension in the Chinese Han population.

METHODS

Seven single nucleotide polymorphisms (SNPs) of the PPARG gene and five SNPs of the AGTR1 gene were selected and genotyped in 1591 unrelated Chinese Han adults. The SNPAssoc package of R was used to analyze the associations between the selected SNPs and hypertension. The potential gene-gene interactions between PPARG and AGTR1 genes were tested by model-based multifactor dimensionality reduction (MB-MDR).

RESULTS

The frequencies of the C allele of rs3856806 and the G allele of rs13433696 in the PPARG gene were significantly lower in hypertensive subjects, whereas the A allele of rs9817428 in the PPARG gene was much higher in hypertensives. In addition, individuals with T allele of rs2933249 in the AGTR1 gene displayed a significantly decreased risk of hypertension. MB-MDR analyses suggested that the two-locus model (rs9817428 and rs2933249) and the three-locus model (rs9817428, rs3856806, and rs2933249) were significantly associated with a decreased risk of hypertension. Moreover, among the eight SNPs not individually associated with hypertension (rs12631819, rs2920502, rs1175543, and rs2972164 in the PPARG gene, and rs2638360, rs1492100, rs5182, and rs275646 in the AGTR1 gene), the two-locus model involving rs12631819 and rs5182 demonstrated increased susceptibility to hypertension, and the five-locus model involving rs12631819, rs2920502, rs2972164, rs5182, and rs2638360 demonstrated a significantly decreased risk of hypertension.

CONCLUSION

Polymorphisms in both the PPARG and AGTR1 genes were found to be significantly associated with hypertension. Moreover, there were significant gene-gene interactions identified between the PPARG and AGTR1 genes in relation to hypertension susceptibility in the Chinese Han population.

Association between CYP3A5 genotypes with hypertension in Chinese Han population: A case-control study

BACKGROUND

The association of CYP3A5 gene polymorphisms with hypertension in the Chinese population is unknown. We explored the association between the CYP3A5 (rs776746) gene and hypertension in the Chinese Han population.

METHODS

Using a case-control design, 340 cases and 254 controls were enrolled from the Third Affiliated Hospital of South Medical University between July and December of 2015. We used a standardized questionnaire to collect data regarding age, sex, smoking, drinking, family history of hypertension, and physical exercise. Height and weight were measured, and the body mass index (BMI) was calculated by weight/height². Blood pressure was measured three times after 5 min of rest with at least 15 s between measurements, and the mean was considered the final BP. A Clinical examination was conducted.

RESULTS

A total of 594 participants, including 340 cases and 254 controls, were entered into the analyses. The genotype frequencies of the CYP3A5 G>A polymorphism did not deviate from the Hardy-Weinberg equilibrium. The genotype frequencies among the cases were 38.8% (GA, 132 individuals), 42.9% (GG, 146 individuals), and 18.2% (AA, 62 individuals). The differences in genotype between the cases and the controls were statistically significant. The AA genotype was associated with an elevated risk of hypertension after adjusting for potential confounders in Model 2. There was no interaction between smoking and the CYP3A5 genotype, while the interaction between drinking and the CYP3A5 genotype was significant.

CONCLUSION

The CYP3A5 gene may be associated with the risk of hypertension in the Chinese Han population, and this effect may be exacerbated by drinking.

Identification of interactions using model-based multifactor dimensionality reduction

BACKGROUND

Common complex traits may involve multiple genetic and environmental factors and their interactions. Many methods have been proposed to identify these interaction effects, among them several machine learning and data mining methods. These are attractive for identifying interactions because they do not rely on specific genetic model assumptions. To handle the computational burden arising from an exhaustive search, including all possible combinations of factors, filter methods try to select promising factors in advance.

METHODS

Model-based multifactor dimensionality reduction (MB-MDR), a semiparametric machine learning method allowing adjustment for confounding variables and lower level effects, is applied to Genetic Analysis Workshop 19 (GAW19) data to identify interaction effects on different traits. Several filtering methods based on the nearest neighbor algorithm are assessed in terms of compatibility with MB-MDR.

RESULTS

Single nucleotide polymorphism (SNP) rs859400 shows a significant interaction effect (corrected p value <0.05) with age on systolic blood pressure (SBP). We identified 23 SNP-SNP interaction effects on hypertension status (HS), 42 interaction effects on SBP, and 26 interaction effects on diastolic blood pressure (DBP). Several of these SNPs are in strong linkage disequilibrium (LD). Three of the interaction effects on HS are identified in filtered subsets.

CONCLUSIONS

The considered filtering methods seem not to be appropriate to use with MB-MDR. LD pruning is further quality control to be incorporated, which can reduce the combinatorial burden by removing redundant SNPs.

Machine learning and data mining in complex genomic data--a review on the lessons learned in Genetic Analysis Workshop 19

In the analysis of current genomic data, application of machine learning and data mining techniques has become more attractive given the rising complexity of the projects. As part of the Genetic Analysis Workshop 19, approaches from this domain were explored, mostly motivated from two starting points. First, assuming an underlying structure in the genomic data, data mining might identify this and thus improve downstream association analyses. Second, computational methods for machine learning need to be developed further to efficiently deal with the current wealth of data.In the course of discussing results and experiences from the machine learning and data mining approaches, six common messages were extracted. These depict the current state of these approaches in the application to complex genomic data. Although some challenges remain for future studies, important forward steps were taken in the integration of different data types and the evaluation of the evidence. Mining the data for underlying genetic or phenotypic structure and using this information in subsequent analyses proved to be extremely helpful and is likely to become of even greater use with more complex data sets.

gammaMAXT: a fast multiple-testing correction algorithm

BACKGROUND

The purpose of the MaxT algorithm is to provide a significance test algorithm that controls the family-wise error rate (FWER) during simultaneous hypothesis testing. However, the requirements in terms of computing time and memory of this procedure are proportional to the number of investigated hypotheses. The memory issue has been solved in 2013 by Van Lishout's implementation of MaxT, which makes the memory usage independent from the size of the dataset. This algorithm is implemented in MBMDR-3.0.3, a software that is able to identify genetic interactions, for a variety of SNP-SNP based epistasis models effectively. On the other hand, that implementation turned out to be less suitable for genome-wide interaction analysis studies, due to the prohibitive computational burden.

RESULTS

In this work we introduce gammaMAXT, a novel implementation of the maxT algorithm for multiple testing correction. The algorithm was implemented in software MBMDR-4.2.2, as part of the MB-MDR framework to screen for SNP-SNP, SNP-environment or SNP-SNP-environment interactions at a genome-wide level. We show that, in the absence of interaction effects, test-statistics produced by the MB-MDR methodology follow a mixture distribution with a point mass at zero and a shifted gamma distribution for the top 10 % of the strictly positive values. We show that the gammaMAXT algorithm has a power comparable to MaxT and maintains FWER, but requires less computational resources and time. We analyze a dataset composed of 10(6) SNPs and 1000 individuals within one day on a 256-core computer cluster. The same analysis would take about 10(4) times longer with MBMDR-3.0.3.

CONCLUSIONS

These results are promising for future GWAIs. However, the proposed gammaMAXT algorithm offers a general significance assessment and multiple testing approach, applicable to any context that requires performing hundreds of thousands of tests. It offers new perspectives for fast and efficient permutation-based significance assessment in large-scale (integrated) omics studies.

Prioritizing hypothesis tests for high throughput data

MOTIVATION

The advent of high throughput data has led to a massive increase in the number of hypothesis tests conducted in many types of biological studies and a concomitant increase in stringency of significance thresholds. Filtering methods, which use independent information to eliminate less promising tests and thus reduce multiple testing, have been widely and successfully applied. However, key questions remain about how to best apply them: When is filtering beneficial and when is it detrimental? How good does the independent information need to be in order for filtering to be effective? How should one choose the filter cutoff that separates tests that pass the filter from those that don't?

RESULT

We quantify the effect of the quality of the filter information, the filter cutoff and other factors on the effectiveness of the filter and show a number of results: If the filter has a high probability (e.g. 70%) of ranking true positive features highly (e.g. top 10%), then filtering can lead to dramatic increase (e.g. 10-fold) in discovery probability when there is high redundancy in information between hypothesis tests. Filtering is less effective when there is low redundancy between hypothesis tests and its benefit decreases rapidly as the quality of the filter information decreases. Furthermore, the outcome is highly dependent on the choice of filter cutoff. Choosing the cutoff without reference to the data will often lead to a large loss in discovery probability. However, naïve optimization of the cutoff using the data will lead to inflated type I error. We introduce a data-based method for choosing the cutoff that maintains control of the family-wise error rate via a correction factor to the significance threshold. Application of this approach offers as much as a several-fold advantage in discovery probability relative to no filtering, while maintaining type I error control. We also introduce a closely related method of P-value weighting that further improves performance.

AVAILABILITY AND IMPLEMENTATION

R code for calculating the correction factor is available at http://www.stat.uga.edu/people/faculty/paul-schliekelman

CONTACT

pdschlie@stat.uga.edu

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Detecting Genetic Interactions for Quantitative Traits Using m-Spacing Entropy Measure

A number of statistical methods for detecting gene-gene interactions have been developed in genetic association studies with binary traits. However, many phenotype measures are intrinsically quantitative and categorizing continuous traits may not always be straightforward and meaningful. Association of gene-gene interactions with an observed distribution of such phenotypes needs to be investigated directly without categorization. Information gain based on entropy measure has previously been successful in identifying genetic associations with binary traits. We extend the usefulness of this information gain by proposing a nonparametric evaluation method of conditional entropy of a quantitative phenotype associated with a given genotype. Hence, the information gain can be obtained for any phenotype distribution. Because any functional form, such as Gaussian, is not assumed for the entire distribution of a trait or a given genotype, this method is expected to be robust enough to be applied to any phenotypic association data. Here, we show its use to successfully identify the main effect, as well as the genetic interactions, associated with a quantitative trait.

An efficiency analysis of high-order combinations of gene-gene interactions using multifactor-dimensionality reduction

Background

Multifactor dimensionality reduction (MDR) is widely used to analyze interactions of genes to determine the complex relationship between diseases and polymorphisms in humans. However, the astronomical number of high-order combinations makes MDR a highly time-consuming process which can be difficult to implement for multiple tests to identify more complex interactions between genes. This study proposes a new framework, named fast MDR (FMDR), which is a greedy search strategy based on the joint effect property.

Results

Six models with different minor allele frequencies (MAFs) and different sample sizes were used to generate the six simulation data sets. A real data set was obtained from the mitochondrial D-loop of chronic dialysis patients. Comparison of results from the simulation data and real data sets showed that FMDR identified significant gene–gene interaction with less computational complexity than the MDR in high-order interaction analysis.

Conclusion

FMDR improves the MDR difficulties associated with the computational loading of high-order SNPs and can be used to evaluate the relative effects of each individual SNP on disease susceptibility. FMDR is freely available at http://bioinfo.kmu.edu.tw/FMDR.rar.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1717-8) contains supplementary material, which is available to authorized users.

A roadmap to multifactor dimensionality reduction methods

Complex diseases are defined to be determined by multiple genetic and environmental factors alone as well as in interactions. To analyze interactions in genetic data, many statistical methods have been suggested, with most of them relying on statistical regression models. Given the known limitations of classical methods, approaches from the machine-learning community have also become attractive. From this latter family, a fast-growing collection of methods emerged that are based on the Multifactor Dimensionality Reduction (MDR) approach. Since its first introduction, MDR has enjoyed great popularity in applications and has been extended and modified multiple times. Based on a literature search, we here provide a systematic and comprehensive overview of these suggested methods. The methods are described in detail, and the availability of implementations is listed. Most recent approaches offer to deal with large-scale data sets and rare variants, which is why we expect these methods to even gain in popularity.

Investigation of TSPO variants in schizophrenia and antipsychotic treatment outcomes

Aim: TSPO is a neuroinflammatory biomarker and emerging therapeutic target in psychiatric disorders. We evaluated whether TSPO polymorphisms contribute to interindividual variability in schizophrenia, antipsychotic efficacy and antipsychotic-induced weight gain. Patients & methods: We analyzed TSPO polymorphisms in 670 schizophrenia cases and 775 healthy controls. Gene–gene interactions between TSPO and other mitochondrial membrane protein-encoding genes (VDAC1 and ANT1) were explored. Positive findings were evaluated in two independent samples (Munich, n = 300; RUPP, n = 119). Results: TSPO rs6971 was independently associated with antipsychotic-induced weight gain in the discovery (puncor = 0.04) and RUPP samples (p = 3.00 × 10-3), and interacted with ANT1 rs10024068 in the discovery (p = 1.15 × 10-3) and RUPP samples (p = 2.76 × 10-4). Conclusion: Our findings highlight TSPO as a candidate for future investigations of antipsychotic-induced weight gain, and support the involvement of mitochondrial membrane components in this serious treatment side effect.

Original submitted 20 August 2014; Revision submitted 3 November 2014

Polymorphisms in circadian genes, night work and breast cancer: results from the GENICA study

OBJECTIVES

The role of genetic variants and environmental factors in breast cancer etiology has been intensively studied in the last decades. Gene-environment interactions are now increasingly being investigated to gain more insights into the development of breast cancer, specific subtypes, and therapeutics. Recently, night shift work that involves circadian disruption has gained rising interest as a potential non-genetic breast cancer risk factor. Here, we analyzed genetic polymorphisms in genes of cellular clocks, melatonin biosynthesis and signaling and their association with breast cancer as well as gene-gene and gene-night work interactions in a German case-control study on breast cancer.

METHODS

GENICA is a population-based case-control study on breast cancer conducted in the Greater Region of Bonn. Associations between seven polymorphisms in circadian genes (CLOCK, NPAS2, ARTNL, PER2 and CRY2), genes of melatonin biosynthesis and signaling (AANAT and MTNR1B) and breast cancer were analyzed with conditional logistic regression models, adjusted for potential confounders for 1022 cases and 1014 controls. Detailed shift-work information was documented for 857 breast cancer cases and 892 controls. Gene-gene and gene-shiftwork interactions were analyzed using model-based multifactor dimensionality reduction (mbMDR).

RESULTS

For combined heterozygotes and rare homozygotes a slightly elevated breast cancer risk was found for rs8150 in gene AANAT (OR 1.17; 95% CI 1.01-1.36), and a reduced risk for rs3816358 in gene ARNTL (OR 0.82; 95% CI 0.69-0.97) in the complete study population. In the subgroup of shift workers, rare homozygotes for rs10462028 in the CLOCK gene had an elevated risk of breast cancer (OR for AA vs. GG: 3.53; 95% CI 1.09-11.42). Shift work and CLOCK gene interactions were observed in the two-way interaction analysis. In addition, gene-shiftwork interactions were detected for MTNR1B with NPAS2 and ARNTL.

CONCLUSIONS

In conclusion, the results of our population-based case-control study support a putative role of the CLOCK gene in the development of breast cancer in shift workers. In addition, higher order interaction analyses suggest a potential relevance of MTNR1B with the key transcriptional factor NPAS2 with ARNTL. Hence, in the context of circadian disruption, multivariable models should be preferred that consider a wide range of polymorphisms, e.g. that may influence chronotype or light sensitivity. The investigation of these interactions in larger studies is needed.

Association of IL33-IL-1 receptor-like 1 (IL1RL1) pathway polymorphisms with wheezing phenotypes and asthma in childhood

BACKGROUND

Genome-wide association studies identified IL33 and IL-1 receptor-like 1 (IL1RL1)/IL18R1 as asthma susceptibility loci. IL33 and IL1RL1 constitute a single ligand-receptor pathway.

OBJECTIVE

In 2 birth cohorts, the Prevalence and Incidence of Asthma and Mite Allergy (PIAMA) study and Avon Longitudinal Study of Parents and Children (ALSPAC), we analyzed associations of longitudinal wheezing phenotypes and asthma with single nucleotide polymorphisms (SNPs) of 8 genes encoding IL-33, IL1RL1, its coreceptor IL1RAcP, its adaptors myeloid differentiation primary response gene 88 (MyD88) and Toll-IL-11 receptor domain containing adaptor protein (TIRAP), and the downstream IL-1 receptor-associated kinase 1, IL-1 receptor-associated kinase 4, and TNF receptor-associated factor 6 (TRAF6). Furthermore, we investigated whether SNPs in this pathway show replicable evidence of gene-gene interaction.

METHODS

Ninety-four SNPs were investigated in 2007 children in the PIAMA study and 7247 children in ALSPAC. Associations with wheezing phenotypes and asthma at 8 years of age were analyzed in each cohort and subsequently meta-analyzed. Gene-gene interactions were assessed through model-based multifactor dimensionality reduction in the PIAMA study, and gene-gene interactions of 10 SNP pairs were further evaluated.

RESULTS

Intermediate-onset wheeze was associated with SNPs in several genes in the IL33-IL1RL1 pathway after applying multiple testing correction in the meta-analysis: 2 IL33 SNPs (rs4742170 and rs7037276), 1 IL-1 receptor accessory protein (IL1RAP) SNP (rs10513854), and 1 TRAF6 SNP (rs5030411). Late-onset wheeze was associated with 2 IL1RL1 SNPs (rs10208293 and rs13424006), and persistent wheeze was associated with 1 IL33 SNP (rs1342326) and 1 IL1RAP SNP (rs9290936). IL33 and IL1RL1 SNPs were nominally associated with asthma. Three SNP pairs showed interaction for asthma in the PIAMA study but not in ALSPAC.

CONCLUSIONS

IL33-IL1RL1 pathway polymorphisms are associated with asthma and specific wheezing phenotypes; that is, most SNPs are associated with intermediate-onset wheeze, a phenotype closely associated with sensitization. We speculate that IL33-IL1RL1 pathway polymorphisms affect development of wheeze and subsequent asthma through sensitization in early childhood.

Multivariate generalized multifactor dimensionality reduction to detect gene-gene interactions

Background

Recently, one of the greatest challenges in genome-wide association studies is to detect gene-gene and/or gene-environment interactions for common complex human diseases. Ritchie et al. (2001) proposed multifactor dimensionality reduction (MDR) method for interaction analysis. MDR is a combinatorial approach to reduce multi-locus genotypes into high-risk and low-risk groups. Although MDR has been widely used for case-control studies with binary phenotypes, several extensions have been proposed. One of these methods, a generalized MDR (GMDR) proposed by Lou et al. (2007), allows adjusting for covariates and applying to both dichotomous and continuous phenotypes. GMDR uses the residual score of a generalized linear model of phenotypes to assign either high-risk or low-risk group, while MDR uses the ratio of cases to controls.

Methods

In this study, we propose multivariate GMDR, an extension of GMDR for multivariate phenotypes. Jointly analysing correlated multivariate phenotypes may have more power to detect susceptible genes and gene-gene interactions. We construct generalized estimating equations (GEE) with multivariate phenotypes to extend generalized linear models. Using the score vectors from GEE we discriminate high-risk from low-risk groups. We applied the multivariate GMDR method to the blood pressure data of the 7,546 subjects from the Korean Association Resource study: systolic blood pressure (SBP) and diastolic blood pressure (DBP). We compare the results of multivariate GMDR for SBP and DBP to the results from separate univariate GMDR for SBP and DBP, respectively. We also applied the multivariate GMDR method to the repeatedly measured hypertension status from 5,466 subjects and compared its result with those of univariate GMDR at each time point.

Results

Results from the univariate GMDR and multivariate GMDR in two-locus model with both blood pressures and hypertension phenotypes indicate best combinations of SNPs whose interaction has significant association with risk for high blood pressures or hypertension. Although the test balanced accuracy (BA) of multivariate analysis was not always greater than that of univariate analysis, the multivariate BAs were more stable with smaller standard deviations.

Conclusions

In this study, we have developed multivariate GMDR method using GEE approach. It is useful to use multivariate GMDR with correlated multiple phenotypes of interests.

MDR-ER: balancing functions for adjusting the ratio in risk classes and classification errors for imbalanced cases and controls using multifactor-dimensionality reduction

Background

Determining the complex relationship between diseases, polymorphisms in human genes and environmental factors is challenging. Multifactor dimensionality reduction (MDR) has proven capable of effectively detecting statistical patterns of epistasis. However, MDR has its weakness in accurately assigning multi-locus genotypes to either high-risk and low-risk groups, and does generally not provide accurate error rates when the case and control data sets are imbalanced. Consequently, results for classification error rates and odds ratios (OR) may provide surprising values in that the true positive (TP) value is often small.

Methodology/Principal Findings

To address this problem, we introduce a classifier function based on the ratio between the percentage of cases in case data and the percentage of controls in control data to improve MDR (MDR-ER) for multi-locus genotypes to be classified correctly into high-risk and low-risk groups. In this study, a real data set with different ratios of cases to controls (1∶4) was obtained from the mitochondrial D-loop of chronic dialysis patients in order to test MDR-ER. The TP and TN values were collected from all tests to analyze to what degree MDR-ER performed better than MDR.

Conclusions/Significance

Results showed that MDR-ER can be successfully used to detect the complex associations in imbalanced data sets.

An epistatic interaction between the PAX8 and STK17B genes in papillary thyroid cancer susceptibility

Papillary Thyroid Cancer (PTC) is a heterogeneous and complex disease; susceptibility to PTC is influenced by the joint effects of multiple common, low-penetrance genes, although relatively few have been identified to date. Here we applied a rigorous combined approach to assess both the individual and epistatic contributions of genetic factors to PTC susceptibility, based on one of the largest series of thyroid cancer cases described to date. In addition to identifying the involvement of TSHR variation in classic PTC, our pioneer study of epistasis revealed a significant interaction between variants in STK17B and PAX8. The interaction was detected by MD-MBR (p = 0.00010) and confirmed by other methods, and then replicated in a second independent series of patients (MD-MBR p = 0.017). Furthermore, we demonstrated an inverse correlation between expression of PAX8 and STK17B in a set of cell lines derived from human thyroid carcinomas. Overall, our work sheds additional light on the genetic basis of thyroid cancer susceptibility, and suggests a new direction for the exploration of the inherited genetic contribution to disease using association studies.

A modified entropy-based approach for identifying gene-gene interactions in case-control study

Gene-gene interactions may play an important role in the genetics of a complex disease. Detection and characterization of gene-gene interactions is a challenging issue that has stimulated the development of various statistical methods to address it. In this study, we introduce a method to measure gene interactions using entropy-based statistics from a contingency table of trait and genotype combinations. We also developed an exploration procedure by using graphs. We propose a standardized relative information gain (RIG) measure to evaluate the interactions between single nucleotide polymorphism (SNP) combinations. To identify the k^th order interactions, contingency tables of trait and genotype combinations of k SNPs are constructed, with which RIGs are calculated. The RIGs are standardized using the mean and standard deviation from the permuted datasets. SNP combinations yielding high standardized RIG are chosen for gene-gene interactions. Detection of high-order interactions and comparison of interaction strengths between different orders are made possible by using standardized RIG. We have applied the proposed standardized entropy-based method to two types of data sets from a simulation study and a real genetic association study. We have compared our method and the multifactor dimensionality reduction (MDR) method through power analysis of eight different genetic models with varying penetrance rates, number of SNPs, and sample sizes. Our method shows successful identification of genetic associations and gene-gene interactions both in simulation and real genetic data. Simulation results suggest that the proposed entropy-based method is better able to detect high-order interactions and is superior to the MDR method in most cases. The proposed method is well suited for detecting interactions without main effects as well as for models including main effects.

Gene-gene interaction analysis for the survival phenotype based on the Cox model

Motivation: For the past few decades, many statistical methods in genome-wide association studies (GWAS) have been developed to identify SNP–SNP interactions for case-control studies. However, there has been less work for prospective cohort studies, involving the survival time. Recently, Gui et al. (2011) proposed a novel method, called Surv-MDR, for detecting gene–gene interactions associated with survival time. Surv-MDR is an extension of the multifactor dimensionality reduction (MDR) method to the survival phenotype by using the log-rank test for defining a binary attribute. However, the Surv-MDR method has some drawbacks in the sense that it needs more intensive computations and does not allow for a covariate adjustment. In this article, we propose a new approach, called Cox-MDR, which is an extension of the generalized multifactor dimensionality reduction (GMDR) to the survival phenotype by using a martingale residual as a score to classify multi-level genotypes as high- and low-risk groups. The advantages of Cox-MDR over Surv-MDR are to allow for the effects of discrete and quantitative covariates in the frame of Cox regression model and to require less computation than Surv-MDR.

Results: Through simulation studies, we compared the power of Cox-MDR with those of Surv-MDR and Cox regression model for various heritability and minor allele frequency combinations without and with adjusting for covariate. We found that Cox-MDR and Cox regression model perform better than Surv-MDR for low minor allele frequency of 0.2, but Surv-MDR has high power for minor allele frequency of 0.4. However, when the effect of covariate is adjusted for, Cox-MDR and Cox regression model perform much better than Surv-MDR. We also compared the performance of Cox-MDR and Surv-MDR for a real data of leukemia patients to detect the gene–gene interactions with the survival time.

Contact:leesy@sejong.ac.kr; tspark@snu.ac.kr

Epistatic interaction between BANK1 and BLK in rheumatoid arthritis: results from a large trans-ethnic meta-analysis

BACKGROUND

BANK1 and BLK belong to the pleiotropic autoimmune genes; recently, epistasis between BANK1 and BLK was detected in systemic lupus erythematosus. Although BLK has been reproducibly identified as a risk factor in rheumatoid arthritis (RA), reports are conflicting about the contribution of BANK1 to RA susceptibility. To ascertain the real impact of BANK1 on RA genetic susceptibility, we performed a large meta-analysis including our original data and tested for an epistatic interaction between BANK1 and BLK in RA susceptibility.

PATIENTS AND METHODS

We investigated data for 1,915 RA patients and 1,915 ethnically matched healthy controls genotyped for BANK1 rs10516487 and rs3733197 and BLK rs13277113. The association of each SNP and RA was tested by logistic regression. Multivariate analysis was then used with an interaction term to test for an epistatic interaction between the SNPs in the 2 genes.

RESULTS

None of the SNPs tested individually was significantly associated with RA in the genotyped samples. However, we detected an epistatic interaction between BANK1 rs3733197 and BLK rs13277113 (P(interaction)  = 0.037). In individuals carrying the BLK rs13277113 GG genotype, presence of the BANK1 rs3733197 G allele increased the risk of RA (odds ratio 1.21 [95% confidence interval 1.04-1.41], P = 0.015. Combining our results with those of all other studies in a large trans-ethnic meta-analysis revealed an association of the BANK1 rs3733197 G allele and RA (1.11 [1.02-1.21], P = 0.012).

CONCLUSION

This study confirms BANK1 as an RA susceptibility gene and for the first time provides evidence for epistasis between BANK1 and BLK in RA. Our results illustrate the concept of pleiotropic epistatic interaction, suggesting that BANK1 and BLK might play a role in RA pathogenesis.

An efficient algorithm to perform multiple testing in epistasis screening

BACKGROUND

Research in epistasis or gene-gene interaction detection for human complex traits has grown over the last few years. It has been marked by promising methodological developments, improved translation efforts of statistical epistasis to biological epistasis and attempts to integrate different omics information sources into the epistasis screening to enhance power. The quest for gene-gene interactions poses severe multiple-testing problems. In this context, the maxT algorithm is one technique to control the false-positive rate. However, the memory needed by this algorithm rises linearly with the amount of hypothesis tests. Gene-gene interaction studies will require a memory proportional to the squared number of SNPs. A genome-wide epistasis search would therefore require terabytes of memory. Hence, cache problems are likely to occur, increasing the computation time. In this work we present a new version of maxT, requiring an amount of memory independent from the number of genetic effects to be investigated. This algorithm was implemented in C++ in our epistasis screening software MBMDR-3.0.3. We evaluate the new implementation in terms of memory efficiency and speed using simulated data. The software is illustrated on real-life data for Crohn's disease.

RESULTS

In the case of a binary (affected/unaffected) trait, the parallel workflow of MBMDR-3.0.3 analyzes all gene-gene interactions with a dataset of 100,000 SNPs typed on 1000 individuals within 4 days and 9 hours, using 999 permutations of the trait to assess statistical significance, on a cluster composed of 10 blades, containing each four Quad-Core AMD Opteron(tm) Processor 2352 2.1 GHz. In the case of a continuous trait, a similar run takes 9 days. Our program found 14 SNP-SNP interactions with a multiple-testing corrected p-value of less than 0.05 on real-life Crohn's disease (CD) data.

CONCLUSIONS

Our software is the first implementation of the MB-MDR methodology able to solve large-scale SNP-SNP interactions problems within a few days, without using much memory, while adequately controlling the type I error rates. A new implementation to reach genome-wide epistasis screening is under construction. In the context of Crohn's disease, MBMDR-3.0.3 could identify epistasis involving regions that are well known in the field and could be explained from a biological point of view. This demonstrates the power of our software to find relevant phenotype-genotype higher-order associations.

Interactions between genetic variants in AMH and AMHR2 may modify age at natural menopause

The onset of menopause has important implications on women’s fertility and health. We previously identified genetic variants in genes involved in initial follicle recruitment as potential modifiers of age at natural menopause. The objective of this study was to extend our previous study, by searching for pairwise interactions between tagging single nucleotide polymorphisms (tSNPs) in the 5 genes previously selected (AMH, AMHR2, BMP15, FOXL2, GDF9). We performed a cross-sectional study among 3445 women with a natural menopause participating in the Prospect-EPIC study, a population-based prospective cohort study, initiated between 1993 and 1997. Based on the model-based multifactor dimensionality reduction (MB-MDR) test with a permutation-based maxT correction for multiple testing, we found a statistically significant interaction between rs10407022 in AMH and rs11170547 in AMHR2 (p = 0.019) associated with age at natural menopause. Rs10407022 did not have a statistically significant main effect. However, rs10407022 is an eQTL SNP that has been shown to influence mRNA expression levels in lymphoblastoid cell lines. This study provides additional insights into the genetic background of age at natural menopause and suggests a role of the AMH signaling pathway in the onset of natural menopause. However, these results remain suggestive and replication by independent studies is necessary.