Ranking Single Nucleotide Polymorphisms Research Articles

BLESS: bagged logistic regression for biomarker identification.

The traditional single nucleotide polymorphism (SNP)-wise approach in genome-wide association studies is focused on examining the marginal association between each SNP with the outcome separately and applying multiple testing adjustments to the resulting p-values to reduce false positives. However, the approach suffers a lack of power in identifying biomarkers. We design an ensemble machine learning approach to aggregate results from logistic regression models based on multiple subsamples, which helps to identify biomarkers from high-dimensional genomic data. We use different methods to analyze a genome-wide association study from the Alzheimer's Disease Neuroimaging Initiative. The SNP-wise approach does not identify any significant signal, while our novel approach provides a list of ranked SNPs associated with the cognitive functions of interests.

A genome-wide by PM10 exposure interaction study for blood pressure in Korean adults

Blood pressure (BP) is a typical complex trait, and the genetic susceptibility of individuals to changes in BP induced by air pollution exposure is different. Although interactions of exposure to air pollutants with several candidate genes have been identified, genome-wide interaction studies (GWISs) are needed to understand the association between them with BP. Therefore, we aimed to discover the unique genetic loci for BP that interact with exposure to air pollutants in Korean adults. We ultimately included 1868 participants in the discovery step and classified them into groups of those with low-to-moderate exposure and high exposure to average annual concentration of particulate matter with an aerodynamic diameter ≤ 10 μm (PM10). Because none of the single nucleotide polymorphisms (SNPs) achieved a genome-wide level of significance of pint < 5 × 10–8 for either systolic BP (SBP) or diastolic BP (DBP), we considered the top 10 ranking SNPs for each BP trait. To validate these suggestive SNPs, we finally selected six genetic variants for SBP and five variants for DBP, respectively. In a replication result for SBP, only one SNP (rs12914147) located in an intergenic region of the NR2F2 showed a significant interaction. We also identified several genetic susceptibility loci (e.g., CHST11, TEK, and ITGA1) implicated in candidate mechanisms such as inflammation and oxidative stress in the discovery step, although their interaction effects were not replicated. Our study reports the first GWIS finding to our knowledge, and the association between exposure to PM10 and BP levels may be determined in part by several newly discovered genetic suggestive loci, including NR2F2.

Detecting Weak Signals by Combining Small P-Values in Genetic Association Studies.

We approach the problem of combining top-ranking association statistics or P-values from a new perspective which leads to a remarkably simple and powerful method. Statistical methods, such as the rank truncated product (RTP), have been developed for combining top-ranking associations, and this general strategy proved to be useful in applications for detecting combined effects of multiple disease components. To increase power, these methods aggregate signals across top ranking single nucleotide polymorphisms (SNPs), while adjusting for their total number assessed in a study. Analytic expressions for combined top statistics or P-values tend to be unwieldy, which complicates interpretation and practical implementation and hinders further developments. Here, we propose the augmented rank truncation (ART) method that retains main characteristics of the RTP but is substantially simpler to implement. ART leads to an efficient form of the adaptive algorithm, an approach where the number of top ranking SNPs is varied to optimize power. We illustrate our methods by strengthening previously reported associations of μ-opioid receptor variants with sensitivity to pain.

Utilizing Deep Learning and Genome Wide Association Studies for Epistatic-Driven Preterm Birth Classification in African-American Women

Genome-Wide Association Studies (GWAS) are used to identify statistically significant genetic variants in case-control studies. The main objective is to find single nucleotide polymorphisms (SNPs) that influence a particular phenotype (i.e., disease trait). GWAS typically use a p-value threshold of 5*10-8 to identify highly ranked SNPs. While this approach has proven useful for detecting disease-susceptible SNPs, evidence has shown that many of these are, in fact, false positives. Consequently, there is some ambiguity about the most suitable threshold for claiming genome-wide significance. Many believe that using lower p-values will allow us to investigate the joint epistatic interactions between SNPs and provide better insights into phenotype expression. One example that uses this approach is multifactor dimensionality reduction (MDR), which identifies combinations of SNPs that interact to influence a particular outcome. However, computational complexity is increased exponentially as a function of higher-order combinations making approaches like MDR difficult to implement. Even so, understanding epistatic interactions in complex diseases is a fundamental component for robust genotype-phenotype mapping. In this paper, we propose a novel framework that combines GWAS quality control and logistic regression with deep learning stacked autoencoders to abstract higher-order SNP interactions from large, complex genotyped data for case-control classification tasks in GWAS analysis. We focus on the challenging problem of classifying preterm births which has a strong genetic component with unexplained heritability reportedly between 20-40 percent. A GWAS data set, obtained from dbGap is utilised, which contains predominantly urban low-income African-American women who had normal and preterm deliveries. Epistatic interactions from original SNP sequences were extracted through a deep learning stacked autoencoder model and used to fine-tune a classifier for discriminating between term and preterm births observations. All models are evaluated using standard binary classifier performance metrics. The findings show that important information pertaining to SNPs and epistasis can be extracted from 4,666 raw SNPs generated using logistic regression (p-value = 5*10-3) and used to fit a highly accurate classifier model. The following results (Sen = 0.9562, Spec = 0.8780, Gini = 0.9490, Logloss = 0.5901, AUC = 0.9745, and MSE = 0.2010) were obtained using 50 hidden nodes and (Sen = 0.9289, Spec = 0.9591, Gini = 0.9651, Logloss = 0.3080, AUC = 0.9825, and MSE = 0.0942) using 500 hidden nodes. The results were compared with a Support Vector Machine (SVM), a Random Forest (RF), and a Fishers Linear Discriminant Analysis classifier, which all failed to improve on the deep learning approach.

A potential regulatory region near the EDN3 gene may control both harness racing performance and coat color variation in horses.

The Swedish‐Norwegian Coldblooded trotter and the heavier North‐Swedish draught horse both descend from the North‐Swedish horse, but the Coldblooded trotters have been selected for racing performance while the North‐Swedish draught horse is mainly used for agricultural and forestry work. By comparing the genomes of Coldblooded trotters, North‐Swedish draught horses and Standardbreds for a large number of single‐nucleotide polymorphisms (SNPs), the aim of the study was to identify genetic regions that may be under selection for racing performance. We hypothesized that the selection for racing performance, in combination with unauthorized crossbreeding of Coldblooded trotters and Standardbreds, has created regions in the genome where the Coldblooded trotters and Standardbreds are similar, but differ from the North‐Swedish draught horse. A fixation index (Fst) analysis was performed and sliding window Delta Fst values were calculated across the three breeds. Five windows, where the average Fst between Coldblooded trotters and Standardbreds was low and the average Fst between Coldblooded trotters and North‐Swedish draught horses was high, were selected for further investigation. Associations between the most highly ranked SNPs and harness racing performance were analyzed in 400 raced Coldblooded trotters with race records. One SNP showed a significant association with racing performance, with the CC genotype appearing to be negatively associated. The SNP identified was genotyped in 1915 horses of 18 different breeds. The frequency of the TT genotype was high in breeds typically used for racing and show jumping while the frequency of the CC genotype was high in most pony breeds and draught horses. The closest gene in this region was the Endothelin3 gene (EDN3), a gene mainly involved in melanocyte and enteric neuron development. Both functional genetic and physiological studies are needed to fully understand the possible impacts of the gene on racing performance.

Pooled genome wide association detects association upstream of FCRL3 with Graves' disease.

BackgroundGraves’ disease is an autoimmune thyroid disease of complex inheritance. Multiple genetic susceptibility loci are thought to be involved in Graves’ disease and it is therefore likely that these can be identified by genome wide association studies. This study aimed to determine if a genome wide association study, using a pooling methodology, could detect genomic loci associated with Graves’ disease.ResultsNineteen of the top ranking single nucleotide polymorphisms including HLA-DQA1 and C6orf10, were clustered within the Major Histo-compatibility Complex region on chromosome 6p21, with rs1613056 reaching genome wide significance (p = 5 × 10−8). Technical validation of top ranking non-Major Histo-compatablity complex single nucleotide polymorphisms with individual genotyping in the discovery cohort revealed four single nucleotide polymorphisms with p ≤ 10−4. Rs17676303 on chromosome 1q23.1, located upstream of FCRL3, showed evidence of association with Graves’ disease across the discovery, replication and combined cohorts. A second single nucleotide polymorphism rs9644119 downstream of DPYSL2 showed some evidence of association supported by finding in the replication cohort that warrants further study.ConclusionsPooled genome wide association study identified a genetic variant upstream of FCRL3 as a susceptibility locus for Graves’ disease in addition to those identified in the Major Histo-compatibility Complex. A second locus downstream of DPYSL2 is potentially a novel genetic variant in Graves’ disease that requires further confirmation.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-016-3276-z) contains supplementary material, which is available to authorized users.

Evidence for contribution of common genetic variants within chromosome 8p21.2-8p21.1 to restricted and repetitive behaviors in autism spectrum disorders.

BackgroundRestricted and Repetitive Behaviors (RRB), one of the core symptom categories for Autism Spectrum Disorders (ASD), comprises heterogeneous groups of behaviors. Previous research indicates that there are two or more factors (subcategories) within the RRB domain. In an effort to identify common variants associated with RRB, we have carried out a genome-wide association study (GWAS) using the Autism Genetic Resource Exchange (AGRE) dataset (n = 1,335, all ASD probands of European ancestry) for each identified RRB subcategory, while allowing for comparisons of associated single nucleotide polymorphisms (SNPs) with associated SNPs in the same set of probands analyzed using all the RRB subcategories as phenotypes in a multivariate linear mixed model. The top ranked SNPs were then explored in an independent dataset.ResultsUsing principal component analysis of item scores obtained from Autism Diagnostic Interview-Revised (ADI-R), two distinct subcategories within Restricted and Repetitive Behaviors were identified: Repetitive Sensory Motor (RSM) and Insistence on Sameness (IS). Quantitative RSM and IS scores were subsequently used as phenotypes in a GWAS using the AGRE ASD cohort. Although no associated SNPs with genome-wide significance (P < 5.0E-08) were detected when RSM or IS were analyzed independently, three SNPs approached genome-wide significance when RSM and IS were considered together using multivariate association analysis. These included the top IS-associated SNP, rs62503729 (P-value = 6.48E-08), which is located within chromosome 8p21.2-8p21.1, a locus previously linked to schizophrenia. Notably, all of the most significantly associated SNPs are located in close proximity to STMN4 and PTK2B, genes previously shown to function in neuron development. In addition, several of the top-ranked SNPs showed correlations with STMN4 mRNA expression in adult CEU (Caucasian and European descent) human prefrontal cortex. However, the association signals within chromosome 8p21.2-8p21.1 failed to replicate in an independent sample of 2,588 ASD probands; the insufficient sample size and between-study heterogeneity are possible explanations for the non-replication.ConclusionsOur analysis indicates that RRB in ASD can be represented by two distinct subcategories: RSM and IS. Subsequent univariate and multivariate genome-wide association studies of these RRB subcategories enabled the detection of associated SNPs at 8p21.2-8p21.1. Although these results did not replicate in an independent ASD dataset, genomic features of this region and pathway analysis suggest that common variants in 8p21.2-8p21.1 may contribute to RRB, particularly IS. Together, these observations warrant future studies to elucidate the possible contributions of common variants in 8p21.2-8p21.1 to the etiology of RSM and IS in ASD.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-016-2475-y) contains supplementary material, which is available to authorized users.

GenoWAP: GWAS signal prioritization through integrated analysis of genomic functional annotation.

Genome-wide association study (GWAS) has been a great success in the past decade. However, significant challenges still remain in both identifying new risk loci and interpreting results. Bonferroni-corrected significance level is known to be conservative, leading to insufficient statistical power when the effect size is moderate at risk locus. Complex structure of linkage disequilibrium also makes it challenging to separate causal variants from nonfunctional ones in large haplotype blocks. Under such circumstances, a computational approach that may increase signal replication rate and identify potential functional sites among correlated markers is urgently needed. We describe GenoWAP, a GWAS signal prioritization method that integrates genomic functional annotation and GWAS test statistics. The effectiveness of GenoWAP is demonstrated through its applications to Crohn's disease and schizophrenia using the largest studies available, where highly ranked loci show substantially stronger signals in the whole dataset after prioritization based on a subset of samples. At the single nucleotide polymorphism (SNP) level, top ranked SNPs after prioritization have both higher replication rates and consistently stronger enrichment of eQTLs. Within each risk locus, GenoWAP may be able to distinguish functional sites from groups of correlated SNPs. GenoWAP is freely available on the web at http://genocanyon.med.yale.edu/GenoWAP.

Genome-wide Gene-Asbestos Exposure Interaction Association Study Identifies a Common Susceptibility Variant on 22q13.31 Associated with Lung Cancer Risk.

Occupational asbestos exposure has been found to increase lung cancer risk in epidemiologic studies. We conducted an asbestos exposure-gene interaction analyses among several Caucasian populations who were current or ex-smokers. The discovery phase included 833 Caucasian cases and 739 Caucasian controls, and used a genome-wide association study (GWAS) to identify single-nucleotide polymorphisms (SNP) with gene-asbestos interaction effects. The top ranked SNPs from the discovery phase were replicated within the International Lung and Cancer Consortium (ILCCO). First, in silico replication was conducted in those groups that had GWAS and asbestos exposure data, including 1,548 cases and 1,527 controls. This step was followed by de novo genotyping to replicate the results from the in silico replication, and included 1,539 cases and 1,761 controls. Multiple logistic regression was used to assess the SNP-asbestos exposure interaction effects on lung cancer risk. We observed significantly increased lung cancer risk among MIRLET7BHG (MIRLET7B host gene located at 22q13.31) polymorphisms rs13053856, rs11090910, rs11703832, and rs12170325 heterozygous and homozygous variant allele(s) carriers (P < 5 × 10(-7) by likelihood ratio test; df = 1). Among the heterozygous and homozygous variant allele(s) carriers of polymorphisms rs13053856, rs11090910, rs11703832, and rs12170325, each unit increase in the natural log-transformed asbestos exposure score was associated with age-, sex-, smoking status, and center-adjusted ORs of 1.34 [95% confidence interval (CI), 1.18-1.51], 1.24 (95% CI, 1.14-1.35), 1.28 (95% CI, 1.17-1.40), and 1.26 (95% CI, 1.15-1.38), respectively, for lung cancer risk. Our findings suggest that MIRLET7BHG polymorphisms may be important predictive markers for asbestos exposure-related lung cancer. To our knowledge, our study is the first report using a systematic genome-wide analysis in combination with detailed asbestos exposure data and replication to evaluate asbestos-associated lung cancer risk.

Genome-wide association study for sight-threatening diabetic retinopathy reveals association with genetic variation near the GRB2 gene.

Diabetic retinopathy is a serious complication of diabetes mellitus and can lead to blindness. A genetic component, in addition to traditional risk factors, has been well described although strong genetic factors have not yet been identified. Here, we aimed to identify novel genetic risk factors for sight-threatening diabetic retinopathy using a genome-wide association study. Retinopathy was assessed in white Australians with type 2 diabetes mellitus. Genome-wide association analysis was conducted for comparison of cases of sight-threatening diabetic retinopathy (n = 336) with diabetic controls with no retinopathy (n = 508). Top ranking single nucleotide polymorphisms were typed in a type 2 diabetes replication cohort, a type 1 diabetes cohort and an Indian type 2 cohort. A mouse model of proliferative retinopathy was used to assess differential expression of the nearby candidate gene GRB2 by immunohistochemistry and quantitative western blot. The top ranked variant was rs3805931 with p = 2.66 × 10(-7), but no association was found in the replication cohort. Only rs9896052 (p = 6.55 × 10(-5)) was associated with sight-threatening diabetic retinopathy in both the type 2 (p = 0.035) and the type 1 (p = 0.041) replication cohorts, as well as in the Indian cohort (p = 0.016). The study-wide meta-analysis reached genome-wide significance (p = 4.15 × 10(-8)). The GRB2 gene is located downstream of this variant and a mouse model of retinopathy showed increased GRB2 expression in the retina. Genetic variation near GRB2 on chromosome 17q25.1 is associated with sight-threatening diabetic retinopathy. Several genes in this region are promising candidates and in particular GRB2 is upregulated during retinal stress and neovascularisation.

Transethnic genome-wide meta-analysis for Alzheimer disease

Genetic risk factors for late-onset Alzheimer disease (AD) shared across different ethnic populations are not well understood. Previously published genome-wide association (GWA) studies for AD by the Alzheimer Disease Genetics Consortium (ADGC) including European American (EA; 13,798 cases and 13,364 controls), African American (AA; 1,675 cases and 3,604 controls), Japanese (JPN; 951 cases and 894 controls), and Israeli-Arab (IA; 86 cases and 84 controls) cohorts were meta-analyzed. Transethnic GWA analyses at a single nucleotide polymorphism (SNP)-level and at a gene-level were conducted for the total sample (ALL) and subgroups of APOE ε4-positive (E4P) and ε4-negative (E4N) subjects in a logistic regression model adjusting for age and sex. Top ranked SNPs (P<10−5) and genes (P<10−4) were replicated using 4,259 subjects including 2,467 AD cases and 1,792 cognitively normal controls in an independent Hispanic cohort. Genome-wide significant SNPs (P<5.4x10−8) in transethnic meta-analysis (TE-P) were identified in the GJC1 gene using E4P subjects (best SNP, rs11871429: TE-P=2.7 x10−8). The association with rs11871429 in E4P was primarily from the EA ethnic group (ethnic specific P value: PEA=5.0x10−9, PAA=0.81, PJPN=0.13), noting that all IA subjects were ε4-negative. No other SNPs in genes were genome-wide significant except for previously established AD genes. Genome-wide significant gene-based tests (P<1.6x10−6) in transethnic meta-analysis were observed in three novel loci, SPDYE3 (P=1.3x10−6), PHOSPHO1 (P=3.7x10−7), and C20orf43 (P=1.5x10−6) in ALL subjects, and GJC1 in E4P subjects (P=5.7x10−7). Gene-based association tests in transethnic meta-analysis were also significant with CR1, BIN1, PTK2B, CLU, MS4A6E, PICALM, and ABCA7 in ALL subjects and with KANSL1 in E4N subjects. Suggestive evidence of association (gene-based P<10−5) was also observed with one novel gene (PLEKHG5) and three previously established AD risk genes (CD2AP, ZCWPW1, and CD33) in ALL subjects. Eleven novel top-ranked genes in the discovery sample were nominally significant (P<0.05) in the replication sample (best result, C20orf43, gene-based P: discovery=1.5x10−6, replication=4.0x10−3, meta-analysis=2.1x10−8). This study suggests that the transethnic analysis is a powerful approach to detect novel LOAD susceptibility genes. Future studies are needed to establish the precise roles of these loci in AD pathogenesis.

Genetics of adolescent idiopathic scoliosis in the post-genome-wide association study era.

Recently, analyses of a genome-wide association study (GWAS) conducted in a Japanese population found several common susceptibility variants associated with adolescent idiopathic scoliosis (AIS) (1-3). The top variants identified from this GWAS reside in genomic regions with potentially etiologically relevant genes and have been replicated across multiple populations and ethnicities. As a result, genes involved in biological pathways that are now viewed as promising targets for downstream functional investigation include abnormal somatosensory function ( LBX1 ) (1), delayed ossification of the developing spine ( GPR126 ) (2), skeletal dysplasia ( SOX9 ) (3) and scoliosis associated with syndromic disease ( KCNJ2 ) (3), One of these variants, a single nucleotide polymorphism (SNP) in the 3’-flanking region of the LBX1 gene (rs11190878) (1), has been consistently replicated in several independent populations (4). Other GWASes have also identified novel AIS-associated variants, with some overlap among the top ranked SNPs, including rs11190878 (Albertsen et al ., 2014, ASHG conference abstract) (5).

Identification of susceptibility loci for colorectal cancer in a genome-wide meta-analysis.

To identify common variants influencing colorectal cancer (CRC) risk, we performed a meta-analysis of five genome-wide association studies, comprising 5626 cases and 7817 controls of European descent. We conducted replication of top ranked single nucleotide polymorphisms (SNPs) in additional series totalling 14 037 cases and 15 937 controls, identifying a new CRC risk locus at 10q24.2 [rs1035209; odds ratio (OR) = 1.13, P = 4.54 × 10(-11)]. We also performed meta-analysis of our studies, with previously published data, of several recently purported CRC risk loci. We failed to find convincing evidence for a previously reported genome-wide association at rs11903757 (2q32.3). Of the three additional loci for which evidence of an association in Europeans has been previously described we failed to show an association between rs59336 (12q24.21) and CRC risk. However, for the other two SNPs, our analyses demonstrated new, formally significant associations with CRC. These are rs3217810 intronic in CCND2 (12p13.32; OR = 1.19, P = 2.16 × 10(-10)) and rs10911251 near LAMC1 (1q25.3; OR = 1.09, P = 1.75 × 10(-8)). Additionally, we found some evidence to support a relationship between, rs647161, rs2423297 and rs10774214 and CRC risk originally identified in East Asians in our European datasets. Our findings provide further insights into the genetic and biological basis of inherited genetic susceptibility to CRC.

The application of network label propagation to rank biomarkers in genome-wide Alzheimer's data.

BackgroundRanking and identifying biomarkers that are associated with disease from genome-wide measurements holds significant promise for understanding the genetic basis of common diseases. The large number of single nucleotide polymorphisms (SNPs) in genome-wide studies (GWAS), however, makes this task computationally challenging when the ranking is to be done in a multivariate fashion. This paper evaluates the performance of a multivariate graph-based method called label propagation (LP) that efficiently ranks SNPs in genome-wide data.ResultsThe performance of LP was evaluated on a synthetic dataset and two late onset Alzheimer’s disease (LOAD) genome-wide datasets, and the performance was compared to that of three control methods. The control methods included chi squared, which is a commonly used univariate method, as well as a Relief method called SWRF and a sparse logistic regression (SLR) method, which are both multivariate ranking methods. Performance was measured by evaluating the top-ranked SNPs in terms of classification performance, reproducibility between the two datasets, and prior evidence of being associated with LOAD.On the synthetic data LP performed comparably to the control methods. On GWAS data, LP performed significantly better than chi squared and SWRF in classification performance in the range from 10 to 1000 top-ranked SNPs for both datasets, and not significantly different from SLR. LP also had greater ranking reproducibility than chi squared, SWRF, and SLR. Among the 25 top-ranked SNPs that were identified by LP, there were 14 SNPs in one dataset that had evidence in the literature of being associated with LOAD, and 10 SNPs in the other, which was higher than for the other methods.ConclusionLP performed considerably better in ranking SNPs in two high-dimensional genome-wide datasets when compared to three control methods. It had better performance in the evaluation measures we used, and is computationally efficient to be applied practically to data from genome-wide studies. These results provide support for including LP in the methods that are used to rank SNPs in genome-wide datasets.

Genetic Prediction of Quantitative Lipid Traits: Comparing Shrinkage Models to Gene Scores

Accurate genetic prediction of quantitative traits related to complex disease risk would have potential clinical impact, so investigation of statistical methodology to improve predictive performance is important. We compare a simple approach of polygenic scores using top ranking single nucleotide polymorphisms (SNPs) to a set of shrinkage models, namely Ridge Regression, Lasso and Hyper-Lasso. These penalised regression methods analyse all genotyped SNPs simultaneously, potentially including much larger sets of SNPs in the models, not only those with the smallest P values. We compare the accuracy of these models for predicting low-density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol, two lipid traits of clinical relevance, in the Whitehall II and British Women's Health and Heart Study cohorts, using SNPs from the HumanCVD BeadChip. For gene scores, the most accurate predictions arise from multivariate weighted scores and include only a small number of SNPs, identified as top hits by the HumanCVD BeadChip. Furthermore, there was little benefit from including external results from published sets of SNPs. We found that shrinkage approaches rarely improved significantly on gene score results. Genetic predictive performance is trait specific, depending on the heritability and genetic architecture of the trait, and is limited by the training data sample size. Our results for lipid traits suggest no current benefit of more complex methods over existing gene score methods. Instead, the most important choice for the prediction model is the number of SNPs and selection of the most predictive SNPs to include. However further comparisons, in larger samples and for other phenotypes, would still be of interest.

EpiMiner: A three-stage co-information based method for detecting and visualizing epistatic interactions

Detecting and visualizing nonlinear interactive effects of Single Nucleotide Polymorphisms (SNPs) or epistatic interactions are important topics of signal processing having great mathematical and computational challenges. To address these problems, a three-stage method, epiMiner (epistasis Miner), is proposed based on co-information theory. In screening stage, Co-Information Index (CII) is employed to visualize and rank contributions of individual SNPs to the phenotype, with the number of top ranking SNPs retained to next stage specified by users directly or a support vector machine classifier automatically. In testing stage, co-information and co-information based permutation test are conducted sequentially to search epistatic interactions within the retained SNPs, with the results then ranked by their p-values. For further characterizing broader epistasis landscape, a visualizing stage is designed to dynamically construct epistasis networks by linking pairs of the retained SNPs if their co-information values with respect to the phenotype are stronger than thresholds. The performance of epiMiner is compared with existing methods on a diverse range of simulated data sets containing several epistasis models. Results demonstrate that epiMiner is effective in detecting and visualizing epistatic interactions. In addition, the application of epiMiner on a real Age-related Macular Degeneration (AMD) data set provides several new clues for the exploration of causative factors of AMD. The Matlab version of epiMiner software is available free online at https://sourceforge.net/projects/epiminer/files/.

Predicting Disease Risk Using Bootstrap Ranking and Classification Algorithms

Genome-wide association studies (GWAS) are widely used to search for genetic loci that underlie human disease. Another goal is to predict disease risk for different individuals given their genetic sequence. Such predictions could either be used as a “black box” in order to promote changes in life-style and screening for early diagnosis, or as a model that can be studied to better understand the mechanism of the disease. Current methods for risk prediction typically rank single nucleotide polymorphisms (SNPs) by the p-value of their association with the disease, and use the top-associated SNPs as input to a classification algorithm. However, the predictive power of such methods is relatively poor. To improve the predictive power, we devised BootRank, which uses bootstrapping in order to obtain a robust prioritization of SNPs for use in predictive models. We show that BootRank improves the ability to predict disease risk of unseen individuals in the Wellcome Trust Case Control Consortium (WTCCC) data and results in a more robust set of SNPs and a larger number of enriched pathways being associated with the different diseases. Finally, we show that combining BootRank with seven different classification algorithms improves performance compared to previous studies that used the WTCCC data. Notably, diseases for which BootRank results in the largest improvements were recently shown to have more heritability than previously thought, likely due to contributions from variants with low minimum allele frequency (MAF), suggesting that BootRank can be beneficial in cases where SNPs affecting the disease are poorly tagged or have low MAF. Overall, our results show that improving disease risk prediction from genotypic information may be a tangible goal, with potential implications for personalized disease screening and treatment.

TNFRSF10B polymorphisms and haplotypes associated with increased risk of death in non-small cell lung cancer

Presently, there are few validated biomarkers that can predict survival or treatment response for non-small cell lung cancer (NSCLC) and most are based on tumor markers. Biomarkers based on germ line DNA variations represent a valuable complementary strategy, which could have translational implications by subclassifying patients to tailored, patient-specific treatment. We analyzed single nucleotide polymorphisms (SNPs) in 53 inflammation-related genes among 651 NSCLC patients. Multivariable Cox proportional hazard models, adjusted for lung cancer prognostic factors, were used to assess the association of genotypes and haplotypes with overall survival. Four of the top 15 SNPs associated with survival were located in the TNF-receptor superfamily member 10b (TNFRSF10B) gene. The T-allele of the top ranked SNP (rs11785599) was associated with a 41% increased risk of death (95% confidence interval [CI] = 1.16-1.70) and the other three TNFRSF10B SNPs (rs1047275, rs4460370 and rs883429) exhibited a 35% (95% CI = 1.11-1.65), 29% (95% CI = 1.06-1.57) and 24% (95% CI = 0.99-1.54) increased risk of death, respectively. Haplotype analyses revealed that the most common risk haplotype (TCTT) was associated with a 78% (95% CI = 1.25-2.54) increased risk of death compared with the low-risk haplotype (CGCC). When the data were stratified by treatment, the risk haplotypes exhibited statistically significantly increased risk of death among patients who had surgery only and no statistically significant effects among patients who had surgery and adjuvant chemotherapy. These data suggest that possessing one or more risk alleles in TNFRSF10B is associated with an increased risk of death. Validated germ line biomarkers may have potential important clinical implications by optimizing patient-specific treatment.

Shared genetic factors for age at natural menopause in Iranian and European women

Do differences in heritable genetic factors explain some of the difference in age at natural menopause (ANM) among populations? One single nucleotide polymorphism (SNP)-ANM association (rs16991615) detected in European women was replicated in Iranian women. Genetics plays an important role in ANM, and well-powered genome-wide association studies (GWAS) of ANM performed in European women have discovered many statistically significant SNP-ANM associations. Average ANM varies by ethnicity, and population-specific differences in ANM-associated alleles may in part explain these differences. After quality control procedures, 97 SNPs were analyzed in genotype data of 828 Iranian women who experienced natural menopause. SNP genotyping data were used to perform linear regression analyses with ANM as a quantitative trait. Study participants were drawn from the population-based Tehran Lipid and Glucose Study based in Tehran, Iran. This study was performed between February 2009 and March 2012. Based on an ANM-GWAS literature review, eight SNPs at four loci previously associated with ANM in European women were tested for replication in Iranian women. Linear regression analyses were performed including (n = 828) and excluding (n = 783) women who experience premature ovarian failure (ANM before 40 years of age). In addition, to search for novel population-specific ANM risk alleles, a pool-based GWAS was performed using this collection of Iranian women. Two DNA pools were constructed and compared: an 'early' ANM pool (lower 20(th) percentile of menopause ages, 40-45 years, n = 165) and a 'late' ANM pool (upper 20(th) percentile of menopause ages, 54-65 years, n = 187). Each DNA pool was assayed on four Illumina Human1M-Duo arrays, and allele-based tests of association were used to rank SNPs. One hundred and two highly ranked SNPs were chosen for individual genotyping by Sequenom MassARRAY and association analysis in the Iranian women. One SNP-ANM association previously detected in European women was replicated in Iranian women (rs16991615; β = 1.07, standard error (SE): 0.49, P = 0.02). SNPs at the previously reported 19q13.42 and 6p24.2 loci also approached statistical significance and had consistent SNP effects (magnitude and direction) in Iranian women (rs1172822; β = -0.39, SE: 0.22, P = 0.08; and rs2153157, β = 0.41, SE: 0.21, P = 0.05). We found little evidence for novel SNP-ANM associations in Iranian women; no SNP selected based on the pool-based GWAS achieved genome-wide significance. Due to small sample size this study was powered to reliably detect only moderate-to-large SNP effect sizes. This limited our ability to replicate many of the previously reported SNP-ANM risk alleles and to discover novel SNP-ANM associations' specific to the Iranian population. In performing our pool-based GWAS, a reduction in power was introduced relative to a conventional GWAS. Our results imply that European and Iranian women share ANM-associated genetic variants. Our study was underpowered but for all SNPs tested the direction of the effect was consistent with data from the European study. Therefore, we anticipate that many (if not all) of the ANM-associated SNPs discovered in European women will replicate in Iranian women upon genotyping a sufficient number of women. Our data do not support the hypothesis that population-specific SNP-ANM associations explain population-specific differences in the mean ANM.

Genome-wide SNP analysis of the Systemic Capillary Leak Syndrome (Clarkson disease).

The Systemic Capillary Leak Syndrome (SCLS) is an extremely rare, orphan disease that resembles, and is frequently erroneously diagnosed as, systemic anaphylaxis. The disorder is characterized by repeated, transient, and seemingly unprovoked episodes of hypotensive shock and peripheral edema due to transient endothelial hyperpermeability. SCLS is often accompanied by a monoclonal gammopathy of unknown significance (MGUS). Using Affymetrix Single Nucleotide Polymorphism (SNP) microarrays, we performed the first genome-wide SNP analysis of SCLS in a cohort of 12 disease subjects and 18 controls. Exome capture sequencing was performed on genomic DNA from nine of these patients as validation for the SNP-chip discoveries and de novo data generation. We identified candidate susceptibility loci for SCLS, which included a region flanking CAV3 (3p25.3) as well as SNP clusters in PON1 (7q21.3), PSORS1C1 (6p21.3), and CHCHD3 (7q33). Among the most highly ranked discoveries were gene-associated SNPs in the uncharacterized LOC100130480 gene (rs6417039, rs2004296). Top case-associated SNPs were observed in BTRC (rs12355803, 3rs4436485), ARHGEF18 (rs11668246), CDH13 (rs4782779), and EDG2 (rs12552348), which encode proteins with known or suspected roles in B cell function and/or vascular integrity. 61 SNPs that were significantly associated with SCLS by microarray analysis were also detected and validated by exome deep sequencing. Functional annotation of highly ranked SNPs revealed enrichment of cell projections, cell junctions and adhesion, and molecules containing pleckstrin homology, Ras/Rho regulatory, and immunoglobulin Ig-like C2/fibronectin type III domains, all of which involve mechanistic functions that correlate with the SCLS phenotype. These results highlight SNPs with potential relevance to SCLS.