Differentiating the Cochran‐Armitage Trend Test and Pearson's χ2 Test: Location and Dispersion

In genetic case‐control association studies, a standard practice is to perform the Cochran‐Armitage (CA) trend test with 1 degree‐of‐freedom (d.f.) under the assumption of an additive model. However, when the true genetic model is recessive or near recessive, it is outperformed by Pearson's χ² test with 2 d.f. In this article, we analytically reveal the statistical basis that leads to the phenomenon. First, we show that the CA trend test examines the location shift between the case and control groups, whereas Pearson's χ² test examines both the location and dispersion shifts between the two groups. Second, we show that under the additive model, the effect of location deviation outweighs that of the dispersion deviation and vice versa under a near recessive model. Therefore, Pearson's χ² test is a more robust test than the CA trend test, and it outperforms the latter when the mode of inheritance evolves to the recessive end.     

Decomposing Pearson's χ2 test: A linear regression and its departure from linearity

In case–control genetic association studies, a standard practice is to perform the Cochran‐Armitage (CA) trend test under the assumption of the additive model because of its robustness. We could even identify situations in which it outperformed the analysis model consistent with the underlying inheritance mode. In this article, we analytically reveal the statistical basis that leads to the phenomenon. By elucidating the origin of the CA trend test as a linear regression model, we decompose Pearson's χ²‐test statistic into two components—one is the CA trend test statistic that measures the goodness of fit of the linear regression model, and the other measures the discrepancy between data and the linear regression model. Under this framework, we show that the additive coding scheme, as well as the multiplicative coding scheme, increases the coefficient of determination of the regression model by increasing the spread of data points. We also obtain the conditions under which the CA trend test statistic equals the MAX statistic and Pearson's χ²‐test statistic.     

A Family‐Based Joint Test for Mean and Variance Heterogeneity for Quantitative Traits

Traditional quantitative trait locus (QTL) analysis focuses on identifying loci associated with mean heterogeneity. Recent research has discovered loci associated with phenotype variance heterogeneity (vQTL), which is important in studying genetic association with complex traits, especially for identifying gene–gene and gene–environment interactions. While several tests have been proposed to detect vQTL for unrelated individuals, there are no tests for related individuals, commonly seen in family‐based genetic studies. Here we introduce a likelihood ratio test (LRT) for identifying mean and variance heterogeneity simultaneously or for either effect alone, adjusting for covariates and family relatedness using a linear mixed effect model approach. The LRT test statistic for normally distributed quantitative traits approximately follows χ²‐distributions. To correct for inflated Type I error for non‐normally distributed quantitative traits, we propose a parametric bootstrap‐based LRT that removes the best linear unbiased prediction (BLUP) of family random effect. Simulation studies show that our family‐based test controls Type I error and has good power, while Type I error inflation is observed when family relatedness is ignored. We demonstrate the utility and efficiency gains of the proposed method using data from the Framingham Heart Study to detect loci associated with body mass index (BMI) variability.     

A Likelihood Ratio Test for Genome‐Wide Association under Genetic Heterogeneity

Most existing association tests for genome‐wide association studies (GWASs) fail to account for genetic heterogeneity. Zhou and Pan proposed a binomial‐mixture‐model‐based association test to account for the possible genetic heterogeneity in case‐control studies. The idea is elegant, however, the proposed test requires an expectation‐maximization (EM)‐type iterative algorithm to identify the penalised maximum likelihood estimates and a permutation method to assess p‐values. The intensive computational burden induced by the EM‐algorithm and the permutation becomes prohibitive for direct applications to GWASs. This paper develops a likelihood ratio test (LRT) for GWASs under genetic heterogeneity based on a more general alternative mixture model. In particular, a closed‐form formula for the LRT statistic is derived to avoid the EM‐type iterative numerical evaluation. Moreover, an explicit asymptotic null distribution is also obtained, which avoids using the permutation to obtain p‐values. Thus, the proposed LRT is easy to implement for GWASs. Furthermore, numerical studies demonstrate that the LRT has power advantages over the commonly used Armitage trend test and other existing association tests under genetic heterogeneity. A breast cancer GWAS dataset is used to illustrate the newly proposed LRT.     

Influence of population stratification on population‐based marker‐disease association analysis

Population‐based genetic association analysis may suffer from the failure to control for confounders such as population stratification (PS). There has been extensive study on the influence of PS on candidate gene‐disease association analysis, but much less attention has been paid to its influence on marker‐disease association analysis. In this paper, we focus on the Pearson χ² test and the trend test for marker‐disease association analysis. The mean and variance of the test statistics are derived under presence of PS, so that the power and inflated type I error rate can be evaluated. It is shown that the bias and the variance distortion are not zero in the presence of both PS and penetrance heterogeneity (PH). Unlike candidate gene‐disease association analysis, when PS is present, the bias is not zero no matter whether PH is present or not. This work generalises the published results, where only the fully recessive penetrance model is considered and only the bias is calculated. It is shown that candidate gene‐disease association analysis can be treated as a special case of marker‐disease association analysis. Consequently, our results extend previous studies on candidate gene‐disease association analysis. A simulation study confirms the theoretical findings.     

The expression profile and prognostic value of APE/Ref‐1 and NPM1 in high‐grade serous ovarian adenocarcinoma

To analyze the expression trends and clinical significance of Apurinic/Apyrimidinic Endodeoxyribonuclease 1 (APE1/Ref‐1) and Nucleophosmin (NPM1) proteins in high‐grade serous ovarian adenocarcinoma (HGSC). The expressions of APE1/Ref‐1 and NPM1 proteins in 94 patients with HGSC were determined using the immunohistochemical (IHC) method, and their relationships with clinicopathological features were analyzed by the χ² test or Fisher's exact test. The follow‐up data, Cox proportional hazards univariate and multivariate survival analyses were integrated to evaluate the prognostic factors affecting patients with HGSC. In the normal fallopian tubes, APE1/Ref‐1 and NPM1 protein were mainly distributed in the nuclear. The HGSC experienced changes in the cellular localization of APE1/Ref‐1 and NPM1 protein expressions, which were abnormally expressed in the cytoplasm. The rates of abnormal cytoplasmic expression of APE1/Ref‐1 and NPM1 proteins in 94 patients with HGSC were 69.1% and 73.4%, respectively, which were significantly higher than the normal fallopian tube tissues (p < 0.05). The abnormal cytoplasmic APE1/Ref‐1 and NPM1 are significantly correlated with the lymph node metastasis, chemosensitivity, FIGO staging, and prognosis. The COX multivariate survival analysis showed that the abnormal expression of APE1/Ref‐1 protein, FIGO staging, and lymph node metastasis are independent prognostic factors. Collectively, the abnormal cytoplasmic APE1/Ref‐1 and NPM1 proteins are associated with the oncogenic progression and chemoresistance of HGSC, and predict a poor prognosis.     

Statistical tests for detecting rare variants using variance-stabilising transformations

Next generation sequencing holds great promise for detecting rare variants underlying complex human traits. Due to their extremely low allele frequencies, the normality approximation for a proportion no longer works well. The Fisher's exact method appears to be suitable but it is conservative. We investigate the utility of various variance-stabilising transformations in single marker association analysis on rare variants. Unlike a proportion itself, the variance of the transformed proportions no longer depends on the proportion, making application of such transformations to rare variant association analysis extremely appealing. Simulation studies demonstrate that tests based on such transformations are more powerful than the Fisher's exact test while controlling for type I error rate. Based on theoretical considerations and results from simulation studies, we recommend the test based on the Anscombe transformation over tests with other transformations.     

On Sample Size and Power Calculation for Variant Set‐Based Association Tests

Sample size and power calculations are an important part of designing new sequence‐based association studies. The recently developed SEQPower and SPS programs adopted computationally intensive Monte Carlo simulations to empirically estimate power for a series of variant set association (VSA) test methods including the sequence kernel association test (SKAT). It is desirable to develop methods that can quickly and accurately compute power without intensive Monte Carlo simulations. We will show that the computed power for SKAT based on the existing analytical approach could be inflated especially for small significance levels, which are often of primary interest for large‐scale whole genome and exome sequencing projects. We propose a new χ²‐approximation‐based approach to accurately and efficiently compute sample size and power. In addition, we propose and implement a more accurate “exact” method to compute power, which is more efficient than the Monte Carlo approach though generally involves more computations than the χ² approximation method. The exact approach could produce very accurate results and be used to verify alternative approximation approaches. We implement the proposed methods in publicly available R programs that can be readily adapted when planning sequencing projects.     

Association of MICA and HLA‐B alleles with leprosy in two endemic populations in Brazil

Leprosy is a prevalent disease in Brazil, which ranks as the country with the second highest number of cases in the world. The disease manifests in a spectrum of forms, and genetic differences in the host can help to elucidate the immunopathogenesis. For a better understanding of MICA association with leprosy, we performed a case–control and a family‐based study in two endemic populations in Brazil. MICA and HLA‐B alleles were evaluated in 409 leprosy patients and in 419 healthy contacts by PCR‐SSOP‐Luminex‐based technology. In the familial study, analysis of 46 families was completed by direct sequencing of all exons and 3′/5′untranslated regions, using the Ilumina MiSeq platform. All data were collected between 2006 and 2009. Statistical analysis was performed using the Chi‐square or Fisher's exact test together with a multivariate analysis. Family‐based association was assessed by transmission disequilibrium test (TDT) software FBAT 2.0.4. We found associations between the haplotype MICA*002‐HLA‐B*35 with leprosy in both the per se and the multibacillary (MB) forms when compared to healthy contacts. The MICA allele *008 was associated with the clinical forms of paucibacillary (PB). Additionally, MICA*029 was associated with the clinical forms of MB. The association of MICA*029 allele (MICA‐A4 variant) with the susceptibility to the MB form suggests this variant for the transmembrane domain of the MICA molecule may be a risk factor for leprosy. Two MICA and nine HLA‐B variants were found associated with leprosy per se in the Colônia do Prata population. Linkage disequilibrium analysis revealed perfect linkage disequilibrium (LD) between HLA‐B markers rs2596498 and rs2507992, and high LD (R² = .92) between these and the marker rs2442718. This familial study demonstrates that MICA association signals are not independent from those observed for HLA‐B. Our findings contribute the knowledge pool of the immunogenetics of Hansen's disease and reveals a new association of the MICA*029 allele.     

Hippocampal sclerosis,hippocampal neuron loss patterns and TDP‐43 in the aged population

Hippocampal neuron loss is a common neuropathological feature in old age with various underlying etiologies. Hippocampal sclerosis of aging (HS‐Aging) is neuropathologically characterized by severe CA1 neuronal loss and frequent presence of transactive response DNA‐binding protein of 43 kDa (TDP‐43) aggregations. Its etiology is unclear and currently no standardized approaches to measure HS‐Aging exist. We developed a semi‐quantitative protocol, which captures various hippocampal neuron loss patterns, and compared their occurrence in the context of HS‐Aging, TDP‐43, vascular and tau pathology in 672 brains (TDP‐43 staining n = 642/672, 96%) donated for the population‐based Cambridge City over‐75s Cohort and the Cognitive Function and Ageing Study. HS‐Aging was first evaluated independently from the protocol using the most common criteria defined in literature, and then described in detail through examination of neuron loss patterns and associated pathologies. 34 (5%) cases were identified, with a maximum of five pyramidal neurons in each of over half CA1 fields‐of‐view (x200 magnification), no vascular damage, no neuron loss in CA2‐CA4, but consistent TDP‐43 neuronal solid inclusions and neurites. We also report focal CA1 neuron loss with vascular pathology to affect predominantly CA1 bordering CA2 (Fisher's exact, P = 0.009), whereas neuron loss in the subicular end of CA1 was associated with TDP‐43 inclusions (Fisher's exact, P < 0.001) and high Braak stage (Fisher's exact, P = 0.001). Hippocampal neuron loss in CA4‐CA2 was not associated with TDP‐43. We conclude that hippocampal neuron loss patterns are associated with different etiologies within CA1, and propose that these patterns can be used to form objective criteria for HS‐Aging diagnosis. Finally, based on our results we hypothesize that neuron loss leading to HS‐Aging starts from the subicular end of CA1 when it is associated with TDP‐43 pathology, and that this neurodegenerative process is likely to be significantly more common than “end‐stage” HS‐Aging only.     

Efficient Calculation of P‐value and Power for Quadratic Form Statistics in Multilocus Association Testing

We address the asymptotic and approximate distributions of a large class of test statistics with quadratic forms used in association studies. The statistics of interest take the general form D=X^TA X , where A is a general similarity matrix which may or may not be positive semi‐definite, and X follows the multivariate normal distribution with mean μ and variance matrix Σ, where Σ may or may not be singular. We show that D can be written as a linear combination of independent χ² random variables with a shift. Furthermore, its distribution can be approximated by a χ² or the difference of two χ² distributions. In the setting of association testing, our methods are especially useful in two situations. First, when the required significance level is much smaller than 0.05 such as in a genome scan, the estimation of p‐values using permutation procedures can be challenging. Second, when an EM algorithm is required to infer haplotype frequencies from un‐phased genotype data, the computation can be intensive for a permutation procedure. In either situation, an efficient and accurate estimation procedure would be useful. Our method can be applied to any quadratic form statistic and therefore should be of general interest.     

Genome‐wide association study of theta band event‐related oscillations identifies serotonin receptor gene HTR7 influencing risk of alcohol dependence

Event‐related brain oscillations (EROs) represent highly heritable neuroelectrical correlates of human perception and cognitive performance that exhibit marked deficits in patients with various psychiatric disorders. We report the results of the first genome‐wide association study (GWAS) of an ERO endophenotype—frontal theta ERO evoked by visual oddball targets during P300 response in 1,064 unrelated individuals drawn from a study of alcohol dependence. Forty‐two SNPs of the Illumina HumanHap 1 M microarray were selected from the theta ERO GWAS for replication in family‐based samples (N = 1,095), with four markers revealing nominally significant association. The most significant marker from the two‐stage study is rs4907240 located within ARID protein 5A gene (ARID5A) on chromosome 2q11 (unadjusted, Fisher's combined P = 3.68 × 10^?6). However, the most intriguing association to emerge is with rs7916403 in serotonin receptor gene HTR7 on chromosome 10q23 (combined P = 1.53 × 10^?4), implicating the serotonergic system in the neurophysiological underpinnings of theta EROs. Moreover, promising SNPs were tested for association with diagnoses of alcohol dependence (DSM‐IV), revealing a significant relationship with the HTR7 polymorphism among GWAS case–controls (P = 0.008). Significant recessive genetic effects were also detected for alcohol dependence in both case–control and family‐based samples (P = 0.031 and 0.042, respectively), with the HTR7 risk allele corresponding to theta ERO reductions among homozygotes. These results suggest a role of the serotonergic system in the biological basis of alcohol dependence and underscore the utility of analyzing brain oscillations as a powerful approach to understanding complex genetic psychiatric disorders. © 2010 Wiley‐Liss, Inc.     

The effect of infectious agents on the prevalence of allergies

PurposeThe role of infectious agents in allergy development is ambivalent. On one hand, there are reports of an association between a previous infection (especially a viral respiratory tract infection) and developing hypersensitivity to inhaled allergens, which in turn may increase the risk of developing allergic reactions. On the other hand, there are reports emphasizing a protective effect of a number of infectious agents against allergy development. The aim the study was to find possible associations between a past infectious or parasitic disease and an allergic condition.Material and methodsThe study population was a group of 18,648 subjects. The study, which was a part of the project: ‘Implementation of a System for the Prevention and Early Detection of Allergic Diseases in Poland’, was conducted in 9 selected regions of Poland and used the ECRHS and ISAAC questionnaires adapted for Europe. The following statistical tools were used: Pearson's chi-squared test, Fisher's exact test, and logistic regression.ResultsThis research was an attempt to clear association between a history of measles or viral hepatitis and the likelihood of developing asthma, especially in males (χ² = 5.29; p<0.05). Past parasitic disease showed a clear association with a suspected allergic rhinitis in various groups of patients (differing both in terms of sex and age).ConclusionsA history of some forms of either infectious or parasitic diseases has a measurable effect on the risk of developing allergies.     

Association of human cytomegalovirus DNAaemia and specific granzyme B responses in lung transplant recipients

In lung transplant recipients (LTRs), human cytomegalovirus (HCMV) DNAaemia could be associated with HCMV disease and reduced allograft survival. In the present study we analysed whether or not HCMV‐specific granzyme B (Grz‐B) responses indicating CD8⁺ T cell cytotoxicity exert an impact on HCMV DNAaemia and relate to specific interferon (IFN)‐γ secretion. HCMV‐specific Grz‐B responses were quantitated by enzyme‐linked immunosorbent assay (ELISA) in 70 samples from 39 HCMV seropositive LTRs who were prospectively investigated for HCMV DNA plasma levels and IFN‐γ kinetics using a standardized CD8⁺ T cell assay (QuantiFERON®‐CMV assay). In all LTRs who were protected from HCMV DNAaemia by early and persistent IFN‐γ responses, Grz‐B responses were also detected. In LTRs who developed episodes of HCMV DNAaemia, the Grz‐B responses which were detected prior to viral DNA detection differed significantly in patients who experienced episodes with high (exceeding 1000 copies/ml) and low plasma DNA levels (P = 0·0290, Fisher's exact test). Furthermore, the extent of Grz‐B release prior to viral DNAaemia correlated statistically with the detected levels of IFN‐γ (P < 0·0001, Spearman's rank test). Of note, simultaneous detection of Grz‐B and IFN‐γ secretion was associated significantly with protection from high HCMV DNA plasma levels during the subsequent follow‐up (P = 0·0057, Fisher's exact test), and this association was stronger than for IFN‐γ detection alone. We conclude that, in addition to IFN‐γ responses, Grz‐B secretion by CD8⁺ T cells is essential to control HCMV replication and a simultaneous measurement of IFN‐γ and Grz‐B could contribute to the immune monitoring of LTRs.     

Association studies of the IL-23R gene in autoimmune thyroid disease in the Japanese population

Autoimmune thyroid diseases (AITDs), including Graves' disease (GD) and Hashimoto's thyroiditis (HT), are caused by interplays of genetic factors and environmental triggers. Interleukin-23 and its receptor (IL-23R) guide T cells towards the Th17 phenotype. IL-23R single nucleotide polymorphisms (SNPs) have been shown to be associated with several autoimmune diseases, including Crohn's disease and rheumatoid arthritis, and Graves' ophthalmopathy (GO) in Caucasians. To determine whether variants in the IL-23R gene are associated with AITDs in Japanese, 464 Japanese AITD patients (290 with GD, 174 with HT) and 179 matched Japanese control subjects were genotyped for four SNPs spanning the IL-23R gene. SNPs rs11209026 and rs7530511 were genotyped using TaqMan allelic discrimination assays and SNPs rs2201841 and rs10889677 were genotyped using a fluorescent-based restriction fragment length polymorphism method. Case-control association studies were performed using the χ² and Fisher's exact tests with Yates correction. Of the four SNPs rs11209026 was non-polymorphic in our dataset. The other three SNPs were not associated with GD or GO or HT in our Japanese population. These results suggest that the IL-23R gene is associated with AITDs only in a specific ethnic group.     

The Impact of FOXP3 Polymorphism on the Risk of Allergic Rhinitis: A Meta‐Analysis

Polymorphisms of several genes were reported to be associated with the risk of allergic rhinitis. Here, we first conducted a meta‐analysis to evaluate the potential genetic association between the polymorphisms of the FOXP3 (Forkhead Box P3) gene and the susceptibility to allergic rhinitis. A total of 2671 relevant articles were initially retrieved from the databases of PubMed, Web of Science, Embase, WANFANG/CNKI and Scopus, and six eligible case‐control studies were finally enrolled in our meta‐analysis, according to our strict inclusion/exclusion criteria. Based on the extracted data, Mantel–Haenszel statistic, Cochrane's Q statistic, I² test, subgroup meta‐analysis, Begg's test, Egger's test and sensitivity analysis were performed via Stata/SE 12.0 software. The results of the Mantel–Haenszel statistic regarding rs3761548 showed that no significant difference was observed in the allergic rhinitis case and population‐based control group under the genetic models of A versus C, AA versus CC, CA+AA versus CC, AA versus CC+CA and carrier A versus C (all P‐value of Association Test, P_A > 0.05), apart from CA versus CC (P_A = 0.020). The similar results were obtained in the subgroup analysis of Asian. In addition, we did not obtain the positive result in the meta‐analysis of rs2232365 (all P_A > 0.05). We also excluded the presence of large publication bias through Begg's test and Egger's test, and we confirmed the stability of data by sensitivity analysis. In summary, no significant association between rs3761548, rs2232365 polymorphisms of the FOXP3 gene, and an increased susceptibility to allergic rhinitis was identified based on the published data; however, this conclusion should be confirmed by more studies with increased sample sizes.     

Putative risk alleles for LATE‐NC with hippocampal sclerosis in population‐representative autopsy cohorts

Limbic‐predominant age‐related TAR‐DNA‐binding protein‐43 (TDP‐43) encephalopathy with hippocampal sclerosis pathology (LATE‐NC + HS) is a neurodegenerative disorder characterized by severe hippocampal CA1 neuron loss and TDP‐43‐pathology, leading to cognitive dysfunction and dementia. Polymorphisms in GRN, TMEM106B and ABCC9 are proposed as LATE‐NC + HS risk factors in brain bank collections. To replicate these results in independent population‐representative cohorts, hippocampal sections from brains donated to three such studies (Cambridge City over 75‐Cohort [CC75C], Cognitive Function and Ageing Study [CFAS], and Vantaa 85+ Study) were stained with hematoxylin–eosin (n = 744) and anti‐pTDP‐43 (n = 713), and evaluated for LATE‐NC + HS and TDP‐43 pathology. Single nucleotide polymorphism genotypes in GRN rs5848, TMEM106B rs1990622 and ABCC9 rs704178 were determined. LATE‐NC + HS (n = 58) was significantly associated with the GRN rs5848 genotype (χ²(2) = 20.61, P < 0.001) and T‐allele (χ²(1) = 21.04, P < 0.001), and TMEM106B rs1990622 genotype (Fisher's exact test, P < 0.001) and A‐allele (χ²(1) = 25.75, P < 0.001). No differences in ABCC9 rs704178 genotype or allele frequency were found between LATE‐NC + HS and non‐LATE‐NC + HS neuropathology cases. Dentate gyrus TDP‐43 pathology associated with GRN and TMEM106B variations, but the association with TMEM106B nullified when LATE‐NC + HS cases were excluded. Our results indicate that GRN and TMEM106B are associated with severe loss of CA1 neurons in the aging brain, while ABCC9 was not confirmed as a genetic risk factor for LATE‐NC + HS. The association between TMEM106B and LATE‐NC + HS may be independent of dentate TDP‐43 pathology.     

Elucidation of de novo small insertion/deletion biology with parent‐of‐origin phasing

The mechanisms underlying de novo insertion/deletion (indel) genesis, such as polymerase slippage, have been hypothesized but not well characterized in the human genome. We implemented two methodological improvements, which were leveraged to dissect indel mutagenesis. We assigned de novo variants to parent‐of‐origin (i.e., phasing) with low‐coverage long‐read whole‐genome sequencing, achieving better phasing compared to short‐read sequencing (medians of 84% and 23%, respectively). We then wrote an application programming interface to classify indels into three subtypes according to sequence context. Across three cohorts with different phasing methods (N_trios = 540, all cohorts), we observed that one de novo indel subtype, change in copy count (CCC), was significantly correlated with father's (p = 7.1 × 10^?4) but not mother's (p = .45) age at conception. We replicated this effect in three cohorts without de novo phasing (p_paternal = 1.9 × 10^?9, p_maternal = .61; N_trios = 3,391, all cohorts). Although this is consistent with polymerase slippage during spermatogenesis, the percentage of variance explained by paternal age was low, and we did not observe an association with replication timing. These results suggest that spermatogenesis‐specific events have a minor role in CCC indel mutagenesis, one not observed for other indel subtypes nor for maternal age in general. These results have implications for indel modeling in evolution and disease.     

Antigen‐specific over‐expression of human cartilage glycoprotein 39 on CD4+CD25+ forkhead box protein 3+ regulatory T cells in the generation of glucose‐6‐phosphate isomerase‐induced arthritis

Human cartilage gp‐39 (HC gp‐39) is a well‐known autoantigen in rheumatoid arthritis (RA). However, the exact localization, fluctuation and function of HC gp‐39 in RA are unknown. Therefore, using a glucose‐6‐phosphate isomerase (GPI)‐induced model of arthritis, we investigated these aspects of HC gp‐39 in arthritis. The rise in serum HC gp‐39 levels was detected on the early phase of GPI‐induced arthritis (day 7) and the HC gp‐39 mRNA was increased significantly on splenic CD4⁺T cells on day7, but not on CD11b⁺cells. Moreover, to identify the characterization of HC gp‐39⁺CD4⁺T cells, we assessed the analysis of T helper (Th) subsets. As a result, HC gp‐39 was expressed dominantly in CD4⁺CD25⁺ forkhead box protein 3 (FoxP3)⁺ refulatory T cells (T_reg), but not in Th1, Th2 or Th17 cells. Furthermore, to investigate the effect of HC gp‐39 to CD4⁺T cells, T cell proliferation assay and cytokine production from CD4⁺T cells using recombinant HC gp‐39 was assessed. We found that GPI‐specific T cell proliferation and interferon (IFN)‐γ or interleukin (IL)‐17 production were clearly suppressed by addition of recombinant HC gp‐39. Antigen‐specific over‐expression of HC gp‐39 in splenic CD4⁺CD25⁺ FoxP3⁺ T_reg cells occurs in the induction phase of GPI‐induced arthritis, and addition of recombinant HC gp‐39 suppresses antigen‐specific T‐cell proliferation and cytokine production, suggesting that HC gp‐39 in CD4⁺ T cells might play a regulatory role in arthritis.     

Likelihood Ratio Tests in Rare Variant Detection for Continuous Phenotypes

It is believed that rare variants play an important role in human phenotypes; however, the detection of rare variants is extremely challenging due to their very low minor allele frequency. In this paper, the likelihood ratio test (LRT) and restricted likelihood ratio test (ReLRT) are proposed to test the association of rare variants based on the linear mixed effects model, where a group of rare variants are treated as random effects. Like the sequence kernel association test (SKAT), a state‐of‐the‐art method for rare variant detection, LRT and ReLRT can effectively overcome the problem of directionality of effect inherent in the burden test in practice. By taking full advantage of the spectral decomposition, exact finite sample null distributions for LRT and ReLRT are obtained by simulation. We perform extensive numerical studies to evaluate the performance of LRT and ReLRT, and compare to the burden test, SKAT and SKAT‐O. The simulations have shown that LRT and ReLRT can correctly control the type I error, and the controls are robust to the weights chosen and the number of rare variants under study. LRT and ReLRT behave similarly to the burden test when all the causal rare variants share the same direction of effect, and outperform SKAT across various situations. When both positive and negative effects exist, LRT and ReLRT suffer from few power reductions compared to the other two competing methods; under this case, an additional finding from our simulations is that SKAT‐O is no longer the optimal test, and its power is even lower than that of SKAT. The exome sequencing SNP data from Genetic Analysis Workshop 17 were employed to illustrate the proposed methods, and interesting results are described.