The power of the Transmission Disequilibrium Test in the presence of population stratification

The Transmission Disequilibrium Test (TDT) is a family-based test for association based on the rate of transmission of alleles from heterozygous parents to affected offspring, and has gained popularity as this test preserves the Type I error rate. Population stratification results in a decreased number of heterozygous parents compared to that expected assuming Hardy–Weinberg Equilibrium (Wahlund Effect). We show that population stratification changes the relative proportion of the informative mating types. The decrease in the number of heterozygous parents and the change in the relative proportion of the informative mating types result in significant changes to the sample sizes required to achieve the power desired. We show examples of the changes in sample sizes, and provide an easy method for estimating TDT sample sizes in the presence of population stratification. This method potentially aids in reducing the number of false-negative association studies.     

Differentiating population stratification from genotyping error using family data

Identifying population stratification and genotyping error are important for candidate gene association studies using the Transmission Disequilibrium Test (TDT). Although the TDT retains the prespecified Type I error in the presence of population stratification, the test may have decreased power in the presence of population stratification. Genotyping error can also cause the TDT to have an elevated Type I error. Differentiating population stratification from genotyping error remains a challenge for geneticists. Both genotyping error and population stratification can result in an increase in the observed homozygosity of a sample relative to that expected assuming Hardy-Weinberg Equilibrium (HWE). We show that when family data are available, even if a limited number of markers are genotyped, evaluating the markers that show statistically significant deviation from HWE with the Mating Type Distortion Test (MTDT)--a test based on the mating type distribution--can reliably differentiate genotyping error from population stratification. We simulate data based on several models of genotyping error in previously published literature, and show how this method could be used in practice to assist in differentiating population stratification from systematic genotyping error.     

A Comparison of Case-Control and Family-Based Association Methods: The Example of Sickle-Cell and Malaria

There has been much debate about the relative merits of population‐ and family‐based strategies for testing genetic association, yet there is little empirical data that directly compare the two approaches. Here we compare case‐control and transmission/disequilibrium test (TDT) study designs using a well‐established genetic association, the protective effect of the sickle‐cell trait against severe malaria. We find that the two methods give similar estimates of the level of protection (case‐control odds ratio = 0.10, 95% confidence interval 0.03–0.23; family‐based estimate of the odds ratio = 0.11, 95% confidence interval 0.04–0.25) and similar statistical significance of the result (case‐control: χ²= 41.26, p= 10^?10 , TDT: χ²= 39.06, p= 10^?10 ) when 315 TDT cases are compared to 583 controls. We propose a family plus population control study design, which allows both case‐control and TDT analysis of the cases. This combination is robust against the respective weaknesses of the case‐control and TDT study designs, namely population structure and segregation distortion. The combined study design is especially cost‐effective when cases are difficult to ascertain and, when the case‐control and TDT results agree, offers greater confidence in the result.     

No evidence for a preferential transmission of the methylenetetrahydrofolate reductase 677T allele in families with schizophrenia offspring.

The methylenetetrahydrofolate reductase (MTHFR) 677C > T polymorphism has been associated with an increased risk of schizophrenia in various case-control studies. However, case-control studies are sensitive to population stratification, which is not an issue in family-based studies. We conducted a family-based study comprising 120 families with a schizophrenic family member to explore the association between the parental MTHFR 677C > T polymorphism and schizophrenia risk in offspring. In addition, a meta-analysis was performed using the available studies with data on this subject. Transmission Disequilibrium Test (TDT) analysis showed no preferential transmission of the 677T allele from parents heterozygous for the MTHFR 677C > T polymorphism to schizophrenia offspring (P = 0.27). The genotype relative risks were 1.43 (95% CI: 0.83-2.47) for the 677TT and 1.42 (95% CI: 0.54-3.78) for the 677CT genotype, relative to the 677CC genotype. A meta-analysis using data from family-based studies comprising a total of 416 parent-child triads yielded no evidence implicating the 677T allele in schizophrenia risk (P = 0.58). By applying a log-linear model, we found no asymmetry within parental mating type. Our data provided no evidence that transmission of the MTHFR 677T allele is associated with schizophrenia risk. In addition, we found no evidence that the maternal genotype influences the risk of having schizophrenia offspring substantially.     

Estimation of Parental Effects Using Polygenic Scores

Offspring resemble their parents for both genetic and environmental reasons. Understanding the relative magnitude of these alternatives has long been a core interest in behavioral genetics research, but traditional designs, which compare phenotypic covariances to make inferences about unmeasured genetic and environmental factors, have struggled to disentangle them. Recently, Kong et al. (2018) showed that by correlating offspring phenotypic values with the measured polygenic score of parents’ nontransmitted alleles, one can estimate the effect of “genetic nurture”—a type of passive gene–environment covariation that arises when heritable parental traits directly influence offspring traits. Here, we instantiate this basic idea in a set of causal models that provide novel insights into the estimation of parental influences on offspring. Most importantly, we show how jointly modeling the parental polygenic scores and the offspring phenotypes can provide an unbiased estimate of the variation attributable to the environmental influence of parents on offspring, even when the polygenic score accounts for a small fraction of trait heritability. This model can be further extended to (a) account for the influence of different types of assortative mating, (b) estimate the total variation due to additive genetic effects and their covariance with the familial environment (i.e., the full genetic nurture effect), and (c) model situations where a parental trait influences a different offspring trait. By utilizing structural equation modeling techniques developed for extended twin family designs, our approach provides a general framework for modeling polygenic scores in family studies and allows for various model extensions that can be used to answer old questions about familial influences in new ways.

     

No linkage of P187S polymorphism in NAD(P)H: Quinone oxidoreductase (NQO1/DIA4) and type 1 diabetes in the Danish population

Recent genome screening studies have identified novel regions of possible interest for susceptibility to type 1 diabetes. One of these is a 30‐35 cM region mapping to 16q22‐q24 (D16S515‐D16S520), where also the gene encoding NAD(P)H: quinone oxidoreductase (NQO1) maps. Data has suggested association of a polymorphism (P187S) in the NQO1 gene and type 1 diabetes. NQO1 is involved in protection against oxidative stress, which is likely to be involved in β‐cell destruction. By use of the transmission disequilibrium test (TDT), we analyzed the P187S polymorphism for association to type 1 diabetes in a population‐based sample of 247 Danish nuclear type 1 diabetic families. Random transmission patterns were observed to all affected offspring (p_tdt = 0.82), to index cases (p_tdt = 0.77), as well as to unaffected offspring (p_tdt = 0.93). Hence, the NQO1 polymorphism is not likely to be an etiological mutation underlying the reported linkage of the 16q22‐q24 region. Hum Mutat 14:67–70, 1999. © 1999 Wiley‐Liss, Inc.     

A haptoglobin mating frequency, segregation and ABO blood-group interaction analysis for additional series of families

Data for a further eight series, comprising a total of 2349 families with 5701 tested children, has been analysed for evidence of disturbances in mating frequency, number of offspring and segregation with respect to haptoglobin type. Five of the series were examined also for interaction effects between the ABO blood-group and haptoglobin systems. There was no significant departure from expectation for the frequency of matings in the haptoglobin system with the exception of the mating class Hp 1-1 × 2-1 in series no. 2. However, a comparison of the number of matings in the reciprocal mating classes 1-1 × 2-1 and 2-1 × 1-1 shows a significant excess of 2-1 × 1-1 matings. This excess is reflected even more in the number of children in the 2-1 × 1-1 mating class compared with the number in the reciprocal 1-1 × 2-1 class. There is a significant distortion in the Hp types of offspring in the Hp2-1 × 2-2 mating class for all series combined as well as for series nos. 1 and 8 considered individually, the Hp2-1 offspring being in excess. Series no. 7 shows a deviation also, but in the opposite direction to that in the other series. A comparison of the Hp¹ gene frequencies among children born to parents who are classified as either compatible or incompatible for the ABO blood-group system shows that the Hp¹ frequency is higher for the incompatible class in each of the five series for which information is available, the difference being significant in series nos. 2 and 3.     

An extended transmission/disequilibrium test (TDT) for multi-allele marker loci

The transmission/disequilibrium test (TDT) was recently introduced by Spielman et al. (1993) as a test for linkage and linkage disequilibrium. The test is based on the unequal probability of transmission of two different marker alleles from parents to affected offspring, when the marker locus and the hypothetical disease locus are linked and are in linkage disequilibrium. The probabilities of marker allele transmission to affected offspring conditional on parental genotype have been derived by Ott (1989) for a biallelic marker and a recessive disorder with no phenocopies. Here, we derive the transmission probabilities for a multi-allele marker locus and a generalized single locus disease model in a random sample of affected individuals from a randomly mating population. The form of these transmission probabilities suggests an extension of the TDT to multi-allele marker loci, in which the alternative hypothesis is restricted to take account of the likely pattern of unequal transmission when the recombination fraction is near 0. We show how our extended TDT can be implemented by standard software for logistic regression, although we have also written our own program which is available on request. We have evaluated the approximate power of the test under a range of realistic assumptions, and it appears that the test will often have good power when linkage disequilibrium is strong and if the disease is recessive.     

High Resolution T2 Association Tests of Complex Diseases Based on Family Data

This paper proposes family based Hotelling's T² tests for high resolution linkage disequilibrium (LD) mapping or association studies of complex diseases. Assume that genotype data of multiple markers or haplotype blocks are available for a sample of nuclear families, in which some offspring are affected. Paired Hotelling's T² test statistics are proposed for a high resolution association study using parents as controls for affected offspring, based on two coding methods: haplotype/allele coding and genotype coding. The paired Hotelling's T² tests take not only the correlation between the haplotype blocks or markers into account, but also take the correlation within each parent‐offspring pair into account. The method extends two sample Hotelling's T² test statistics for population case control association studies, which are not valid for family data due to correlation of genetic data among family members. The validity of the proposed method is justified by rigorous mathematical and statistical proof under the large sample theory. The non‐centrality parameter approximations of the test statistics are calculated for power and sample size calculations. From power comparison and type I error calculations, it is shown that the test statistic based on haplotype/allele coding is advantageous over the test statistic of genotype coding. Analysis using multiple markers may provide higher power than single marker analysis. If only one marker is utilized the power of the test statistic based on haplotype/allele coding is nearly identical to that of 1‐TDT. Moreover, a permutation procedure is provided for data analysis. The method is applied to data from a German asthma family study. The results based on the paired Hotelling's T² statistic tests confirm the previous findings. However, the paired Hotelling's T² tests produce much smaller P‐values than those of the previous study. The permutation tests produce similar results to those of the previous study; moreover, additional marker combinations are shown to be significant by permutation tests. The proposed paired Hotelling's T² statistic tests are potentially powerful in mapping complex diseases. A SAS Macro, Hotel_fam.sas, has been written to implement the method for data analysis.     

A Genetic Variant rs1020760at NFKB1 is Associated with Clinical Features of Psoriasis Vulgaris in a Han Chinese Population

Psoriasis vulgaris is a chronic inflammatory skin disease associated with complex genetic susceptibility. Recently, we identified a single‐nucleotide variant rs1020760 at NFKB1 significantly associated with psoriasis in a Han Chinese population in deep analysis of exome and targeted sequencing (P = 1.76 × 10^?8). To investigate the potential association between rs1020760 and phenotypes of psoriasis vulgaris, we performed a genotype–phenotype analysis. A total of 9946 cases and 9906 controls with detailed clinical and demographic information were involved in this study, while the genotype data of rs1020760 was available in the previous targeted sequencing study of psoriasis. Genotype‐based association testing revealed the additive model might provide the best fit for rs1020760 (P = 5.44 × 10^?8). Case‐only analysis showed that the distribution of allele G was significantly different between the cases with and without family history (P_allele = 4.07 × 10^?3,P_genotype = 5.75 × 10^?3). The differences in allele and genotype frequencies were observed between all the subphenotypes and controls except for the genotype frequency of the late onset subgroup, while no difference was found in case‐only analysis for the other two subphenotypes. Rs1020760 was preferentially associated with family history of psoriasis, implying that NFKB1 might not only play important roles in the development of psoriasis, but might also contribute to the special phenotypes of this disease.     

Ancestry Informative Marker Panel to Estimate Population Stratification Using Genome‐wide Human Array

Case‐control studies are a powerful strategy to identify candidate genes in complex diseases. In admixed populations, association studies can be affected by population stratification, leading to spurious genetic associations. Ancestry informative markers (AIMs) can be used to minimise this effect. The aim of this work was to select a set of AIMs to estimate population stratification in a Brazilian case‐control study performed using a genome‐wide array. A total of 345 single nucleotide polymorphism (SNP) AIMs, selected from the Cytoscan HD array and based on previously reported panels, was used to discriminate between European, African, and Amerindian populations. These SNP‐AIMs were used to infer ancestry in systemic lupus erythematosus (SLE) patients (n = 23) and in healthy subjects (n = 110). Moderate population substructure was observed between SLE and control groups (F_{st =} 0.0113). Although patients and controls have shown a major European genomic contribution, significant differences in the European (P = 6.47 × 10^?5) and African (P = 1.14 × 10^?3) ancestries were detected between the two groups. We performed a two‐step validation of the 345 SNP‐AIMs panel estimating the ancestral contributions using a panel of 12 AIMs and approximately 70K SNPs from the array. Evaluation of population substructure in case‐control studies, avoiding spurious genetic associations, can be performed using our panel of 345 SNP‐AIMs.     

Association between the COMT locus and obsessive‐compulsive disorder in females but not males

A polymorphism in the coding region of catechol‐O‐methyltransferase gene (COMT) was previously reported to be associated with obsessive‐compulsive disorder (OCD), particularly in male probands. We attempted to replicate the previous finding using a family‐based genetic design in haplotype relative risk (HRR) and transmission disequilibrium (TDT) analyses. Fifty‐six OCD probands and their parents were genotyped for the COMT locus using established methods. Analysis of allele and genotype frequencies between the proband genotypes and the control (parental nontransmitted) genotypes failed to replicate the previous finding of gender divergence, gave no evidence of overall association, nor was linkage detected by TDT. However, further analysis of the COMT allele frequencies by proband gender gave evidence of a mildly significant association with the low‐activity COMT allele in female probands (P = 0.049), but not in male probands. These findings indicate that COMT may be etiologically relevant to OCD in a gender‐specific manner opposite to that shown in previous studies. © 2001 Wiley‐Liss, Inc.     

Family‐based study of the association of the dopamine D2 receptor gene (DRD2) with habitual smoking

A recent study showed an association between the dopamine D2 receptor gene (DRD2) and smoking. The purpose of this study was to determine if the familial transmission of smoking is linked to variation at the DRD2 locus in a genetically informative sample. Subjects were identified in alcohol treatment centers and their relatives were recruited for study. All subjects were interviewed to assess alcohol dependence, smoking habits, and psychiatric disorders. Two polymorphisms within the DRD2 gene were analyzed, including the TaqIA polymorphism. The sample consisted of 138 nuclear families with at least one offspring with habitual smoking, and analysis was by the transmission disequilibrium test (TDT), which avoids problems due to population stratification. There was no significant difference in the frequency between DRD2 alleles transmitted and not transmitted to habitual smokers. There also was no evidence for unequal transmission of DRD2 alleles for the phenotypes “ever smoker” or comorbid alcohol dependence and habitual smoking. This study does not support linkage of the DRD2 with smoking. Am. J. Med. Genet. 90:299–302, 2000. © 2000 Wiley‐Liss, Inc.     

Metastatic risk stratification of clear cell renal cell carcinoma patients based on genomic aberrations

Prognostic markers for the definition of the individual metastatic risk in renal cell carcinoma are still missing. The aim of our study was to establish a total number of specific aberrations (TNSA) genetic score as a new prognostic test for metastatic risk evaluation. Fluorescence in situ hybridization (FISH) was performed on isolated cell nuclei of 100 ccRCCs (50 M1/50 M0) and 100 FFPE sections (second cohort, 32 M1/68 M0). For each chromosomal region (1q21.3, 7q36.3, 9p21.3p24.1, 20q11.21q13.32) cut‐off values were determined by receiver‐operator curve (ROC)‐curve analysis. TNSA was calculated based on the dichotomized specific CNVs. The prognostic significance of CNVs was proven by Cox and logistic regression. TNSA was the best predictor of metastasis and recurrence free survival in both cohorts. We derived an algorithm for risk stratification by combining TNSA and T‐category, which increased the prognostic accuracy to 87% (specificity = 86%, sensitivity = 88%). This model divides patients into two risk groups with significantly different RFS, CSS, and OS (P = 3.8×10^?5, P = 5×10^?6 and P = 3.57×10^?8 respectively). The genetic risk model was superior to Leibovich score and was able to identify patients with metachronous metastatic spread which were incorrectly classified as “low” or “intermediate risk.” We present a new tool for individual risk stratification by combining genetic alterations with clinico‐pathologic parameters. Interphase FISH proves to be a dependable method for prognostic evaluation in primary tumor tissue on isolated cell nuclei as well as on FFPE sections. Especially in organ‐confined tumors the genetic score seems to be an important tool to identify patients at high risk for metastatic disease.     

Testing informative missingness in genetic studies using case-parent triads

In genetic studies, the transmission/disequilibrium test (TDT) using case-parent triads has gained popularity attributable to its robustness to population admixture. Several extensions have been proposed to accommodate incomplete triads. Some strategies assume that parental genotypes are missing completely at random (MCAR) to insure an unbiased conclusion and some methods allow parental genotypes to be missing informatively, resulting in reduced power when the missing data pattern is indeed MCAR. However, these tests assumed that offspring genotypes were MCAR. Recently, Guo indicated that when offspring genotypes were missing informatively, an occurrence that can be considered as ascertainment bias, inflated type-I error and/or reduced power may occur using the TDT when incomplete triads are excluded. In an effort to avoid an erroneous conclusion, we propose a strategy called testing informative missingness (TIM) that compares conditional distributions of parental genotypes among complete triads and incomplete data with only one parent to examine the missing data pattern. Through computer simulations, TIM has decent power to detect informative missingness and is robust to population admixture. In addition, we illustrate TIM with an application to the Framingham Heart Study.     

Tests for Candidate-Gene Association Using Case-Parents Design

In the case‐parents design for testing candidate‐gene association, the conditional likelihood method based on genotype relative risks has been developed recently. A specific relation of the genotype relative risks is referred to as a genetic model. The efficient score tests have been used when the genetic model is correctly specified under the alternative hypothesis. In practice, however, it is usually not able to specify the genetic model correctly. In the latter situation, tests such as the likelihood ratio test (LRT) and the MAX3 (the maximum of the three score statistics for dominant, additive, and recessive models) have been used. In this paper, we consider the restricted likelihood ratio test (RLRT). For a specific genetic model, simulation results demonstrate that RLRT is asymptotically equivalent to the score test, and both are more powerful than the LRT. When the genetic model cannot be correctly specified, the simulation results show that RLRT is most robust and powerful in the situations we studied. MAX3 is the next most robust and powerful test. The TDT is the easiest statistic to compute, compared to MAX3 and RLRT. When the recessive model can be eliminated, it is also as robust and powerful as RLRT for other genetic models.     

A Marginal Likelihood Model for Family-Based Data

This paper presents a marginal likelihood model for family‐based data based upon the transmission of marker alleles from each heterozygous parent to his/her affected children. The proposed model, extending the maximum‐likelihood‐binomial (MLB) method and the disequilibrium maximum‐likelihood‐binomial (DMLB) method ( Abel et al. 1998 ; Abel & Müller‐Myhsok, 1998 ; Huang & Jiang, 1999 ), is adaptive to linkage disequilibrium (LD) and linkage heterogeneity. Compared with other procedures, the likelihood ratio test (LRT) derived from the proposed model enjoys superior qualities. First, simulations indicate that the power of the LRT is greater than that of the TDT or DMLB in all of our studied scenarios. Second, when we applied the LRT and other tests to a Tourette Syndrome data, the result was data favorable to the use of the LRT. Therefore, we recommend the use of the LRT as an additional linkage test wherever applicable, especially when the amount of LD is uncertain.     

CTLA‐4 49 A/G dimorphism and type 1 diabetes susceptibility: a French case–control study and segregation analysis. Evidence of a maternal effect

Several studies have demonstrated an association of cytotoxic T lymphocyte‐associated molecule 4 (CTLA‐4) (IDDM 12) alanine 17 with type 1 diabetes, but we wished to study the parental effect of CTLA‐4 49 A/G dimorphism in diabetic families. The CTLA‐4 exon 1 polymorphism (49 A/G), HLA‐DRB1 and insulin gene (INS) variable number tandem repeats (VNTR) were analysed in 134 type 1 diabetic patients vs. 273 control subjects. The segregation analysis for transmission was carried out in 70 informative diabetic families using the transmission distortion test (TDT). All genotyping was performed by PCR‐RFLP. CTLA‐4 49 G allele frequency was not increased in diabetic patients compared to controls (41 vs. 38%, not significant). The distribution of GG, AG and AA CTLA‐4 genotypes was similar in the two groups (13, 57 and 30% vs. 11, 54 and 35%, respectively) and was independent of HLA‐DRB1 or INS VNTR polymorphism. The CTLA‐4 49 G allele showed weak distorted transmission to the diabetic offspring, whereas random transmission was observed in unaffected offspring. This distortion is attributable to a maternal effect (71% compared to the 50% expected ratio; tdt = 4.8; P < 0.03). The combined transmission of maternal CTLA‐4 G with HLA‐DRB1*03 (90%; tdt = 6.4; P < 0.01) and VNTR class I (80%; tdt = 5.4; P < 0.02) enhanced the susceptibility effect of each marker separately. We noted a slight CTLA‐4 49 G and HLA‐DRB1*04 distortion of transmission shared in paternal and maternal diabetic meiosis. In non‐diabetic offspring, the CTLA‐4 49 A allele confers a protective effect in the presence of maternal HLA‐DRB1*03 and paternal HLA‐DRB1*04 alleles. Despite the absence of a positive association of the CTLA‐4 49 G allele with type 1 diabetes, our segregation analysis supports the hypothesis of a modulation by CTLA‐4 49 G/A dimorphism of the susceptibility conferred by maternal HLA‐DRB1*03 inheritance. This potential parental effect needs to be confirmed in a larger data set.     

Genome‐wide analyses of psychological resilience in U.S. Army soldiers

Though a growing body of preclinical and translational research is illuminating a biological basis for resilience to stress, little is known about the genetic basis of psychological resilience in humans. We conducted genome‐wide association studies (GWASs) of self‐assessed (by questionnaire) and outcome‐based (incident mental disorders from predeployment to postdeployment) resilience among European (EUR) ancestry soldiers in the Army study to assess risk and resilience in servicemembers. Self‐assessed resilience (N = 11,492) was found to have significant common‐variant heritability (h² = 0.162, se = 0.050, p = 5.37 × 10^?4), and to be significantly negatively genetically correlated with neuroticism (r_g = ?0.388, p = .0092). GWAS results from the EUR soldiers revealed a genome‐wide significant locus on an intergenic region on Chr 4 upstream from doublecortin‐like kinase 2 (DCLK2) (four single nucleotide polymorphisms (SNPs) in LD; top SNP: rs4260523 [p = 5.65 × 10^?9] is an eQTL in frontal cortex), a member of the doublecortin family of kinases that promote survival and regeneration of injured neurons. A second gene, kelch‐like family member 36 (KLHL36) was detected at gene‐wise genome‐wide significance [p = 1.89 × 10^?6]. A polygenic risk score derived from the self‐assessed resilience GWAS was not significantly associated with outcome‐based resilience. In very preliminary results, genome‐wide significant association with outcome‐based resilience was found for one locus (top SNP: rs12580015 [p = 2.37 × 10^?8]) on Chr 12 downstream from solute carrier family 15 member 5 (SLC15A5) in subjects (N = 581) exposed to the highest level of deployment stress. The further study of genetic determinants of resilience has the potential to illuminate the molecular bases of stress‐related psychopathology and point to new avenues for therapeutic intervention.