Quantitative trait locus mapping methods for diversity outbred mice.

Genetic mapping studies in the mouse and other model organisms are used to search for genes underlying complex phenotypes. Traditional genetic mapping studies that employ single-generation crosses have poor mapping resolution and limit discovery to loci that are polymorphic between the two parental strains. Multiparent outbreeding populations address these shortcomings by increasing the density of recombination events and introducing allelic variants from multiple founder strains. However, multiparent crosses present new analytical challenges and require specialized software to take full advantage of these benefits. Each animal in an outbreeding population is genetically unique and must be genotyped using a high-density marker set; regression models for mapping must accommodate multiple founder alleles, and complex breeding designs give rise to polygenic covariance among related animals that must be accounted for in mapping analysis. The Diversity Outbred (DO) mice combine the genetic diversity of eight founder strains in a multigenerational breeding design that has been maintained for >16 generations. The large population size and randomized mating ensure the long-term genetic stability of this population. We present a complete analytical pipeline for genetic mapping in DO mice, including algorithms for probabilistic reconstruction of founder haplotypes from genotyping array intensity data, and mapping methods that accommodate multiple founder haplotypes and account for relatedness among animals. Power analysis suggests that studies with as few as 200 DO mice can detect loci with large effects, but loci that account for <5% of trait variance may require a sample size of up to 1000 animals. The methods described here are implemented in the freely available R package DOQTL.

Genome-wide association studies of maximum number of drinks.

Maximum number of drinks (MaxDrinks) defined as "Maximum number of alcoholic drinks consumed in a 24-h period" is an intermediate phenotype that is closely related to alcohol dependence (AD). Family, twin and adoption studies have shown that the heritability of MaxDrinks is approximately 0.5. We conducted the first genome-wide association (GWA) study and meta-analysis of MaxDrinks as a continuous phenotype. 1059 individuals were from the Collaborative Study on the Genetics of Alcoholism (COGA) sample and 1628 individuals were from the Study of Addiction - Genetics and Environment (SAGE) sample. Family sample with 3137 individuals was from the Australian twin-family study of alcohol use disorder (OZALC). Two population-based Caucasian samples (COGA and SAGE) with 1 million single-nucleotide polymorphisms (SNPs) were used for gene discovery and one family-based Caucasian sample was used for replication. Through meta-analysis we identified 162 SNPs associated with MaxDirnks (p < 10(-4)). The most significant association with MaxDrinks was observed with SNP rs11128951 (p = 4.27 × 10(-8)) near SGOL1 gene at 3p24.3. Furthermore, several SNPs (rs17144687 near DTWD2, rs12108602 near NDST4, and rs2128158 in KCNB2) showed significant associations with MaxDrinks (p < 5 × 10(-7)) in the meta-analysis. Especially, 8 SNPs in DDC gene showed significant associations with MaxDrinks (p < 5 × 10(-7)) in the SAGE sample. Several flanking SNPs in above genes/regions were confirmed in the OZALC family sample. In conclusions, we identified several genes/regions associated with MaxDrinks. These findings can improve the understanding about the pathogenesis of alcohol consumption phenotypes and alcohol-related disorders.

Associations of smoking status and serious psychological distress with chronic obstructive pulmonary disease.

BACKGROUND: Chronic obstructive pulmonary disease (COPD) has been a major public health problem due to its high prevalence, morbidity, and mortality. Smoking is a major risk factor for COPD, while serious psychological distress (SPD) is prevalent among COPD patients. However, no study focusing on the effect of SPD on COPD has been so far conducted, while few studies have focused on the associations of SPD and behavioral factors with COPD by smoking status. OBJECTIVES: This study aimed to examine the associations of SPD and behavioral factors (such as smoking and physical activity) with COPD. MATERIALS AND METHODS: Weighted logistic regression models were used for the analysis of 1,248 cases and 39,995 controls from the 2005 California Health Interview Survey (CHIS). RESULTS: The prevalence of SPD was 10% in cases and 4% in controls, respectively. The percentages of past and current smoking were higher in cases than controls (50% vs. 24% and 27% vs. 15%, respectively). After adjusting for other factors, smoking (OR = 4.56, 95% CI = 3.41-6.11 and OR = 3.24, 95% CI = 2.57-4.08 for current and past smoking, respectively), physical activity (OR = 0.69, 95% CI = 0.55-0.87), obesity (OR = 1.25, 95% CI = 1.03-1.52), older age (OR = 2.86, 95% CI = 2.15-3.82, and OR = 5.97, 95% CI = 4.42-8.08 for middle-aged and elder groups, respectively), SPD (OR = 2.11, 95% CI = 1.47-3.04), employment (OR = 0.62, 95% CI = 0.51-0.76), race (OR = 0.35, 95% CI = 0.23-0.54, OR = 0.59, 95% CI = 0.36-0.97, and OR = 0.47, 95% CI=0.29-0.75 for Latino, Asian, and African American, respectively) and lower federal poverty level (OR=1.89, 95% CI = 1.35-2.63, OR = 1.65, 95% CI = 1.27-2.14, and OR = 1.39, 95% CI = 1.12-1.72 for 0-99% FPL, 100-199% FPL and 200-299% FPL, respectively) were all associated with COPD (P < 0.05). Age group, SPD, race, and employment showed significant interactions with smoking status. Stratified by smoking status, aging was the only risk factor for COPD in the never smoking group; whereas, lack of physical activity, older age, SPD, race, unemployment, and lower federal poverty level were associated with COPD in the smoking groups. CONCLUSIONS: Smoking and aging were major risk factors for COPD, while lack of physical activity and SPD were strongly associated with COPD in the smoking groups.

Common variants in HLA-DRA gene are associated with alcohol dependence in two Caucasian samples.

HLA-DRA gene polymorphisms might play an important role in alcohol dependence (AD). We examined genetic associations of 29 single-nucleotide polymorphisms (SNPs) within the HLA-DRA gene with AD using two Caucasian samples-the Collaborative Study on the Genetics of Alcoholism (COGA) sample (660 AD cases and 400 controls) and the Study of Addiction: Genetics and Environment (SAGE) sample (623 cases and 1,016 controls). Logistic regression analysis using PLINK showed that 16 SNPs were associated with AD in the COGA sample and 13 SNPs were associated with AD in the SAGE sample (p < 0.05). The best novel signal was SNP rs2239803 associated with AD in both samples (p = 0.000817 for the COGA sample and p = 0.0026 for the SAGE sample, respectively) while one flanking SNP rs4935356 also showed strong association in both samples (p = 0.00219 and 0.0026 for the COGA and SAGE samples, respectively). Furthermore, these two SNPs revealed stronger associations in meta-analysis of these two samples (p = 8.97 × 10(-6) and 2.02 × 10(-5) for rs2239803 and rs4935356, respectively). In addition, the G-A haplotype from these two SNPs revealed a significant association with AD in both the COGA and SAGE samples (p = 0.0007 and 0.0019, respectively). These findings highlight the novel associations with HLA-DRA that may play an important role in the etiology of AD.

Genome-wide association study identifies 5q21 and 9p24.1 (KDM4C) loci associated with alcohol withdrawal symptoms.

Several genome-wide association (GWA) studies of alcohol dependence (AD) and alcohol-related phenotypes have been conducted; however, little is known about genetic variants influencing alcohol withdrawal symptoms (AWS). We conducted the first GWA study of AWS using 461 cases of AD with AWS and 408 controls in Caucasian population in the Collaborative Study on the Genetics of Alcoholism (COGA) sample. Logistic regression analysis of AWS as a binary trait, adjusted for age and sex, was performed using PLINK. We identified 51 SNPs associated with AWS with p < 10(-4). The first best signal was rs770182 (p = 3.65 × 10(-6)) at 5q21 near EFNA5 gene which was replicated in the Australian twin-family study of 273 families (p = 0.0172). Furthermore, three SNPs (rs10975990, rs10758821 and rs1407862) within KDM4C gene at 9p24.1 showed p < 10(-4) (p = 7.15 × 10(-6), 2.79 × 10(-5) and 4.93 × 10(-5), respectively) in the COGA sample while one SNP rs12001158 within KDM4C with p = 1.97 × 10(-4) in the COGA sample was replicated in the family sample (p = 0.01). Haplotype analysis further supported the associations of single-marker analyses of KDM4C in the COGA sample. Moreover, two SNPs (rs2046593 and rs10497668) near FSIP2 at 2q32.1 with moderate associations with AWS in the COGA sample (p = 2.66 × 10(-4) and 9.48 × 10(-5), respectively) were replicated in the family sample (p = 0.0013 and 0.0162, respectively). In addition, several SNPs in GABRA1, GABRG1, and GABRG3 were associated with AWS (p < 10(-2)) in the COGA sample. In conclusion, we identified several loci associated with AWS. These findings offer the potential for new insights into the pathogenesis of AD and AWS.

A Bayesian segmentation approach to ascertain copy number variations at the population level.

MOTIVATION: Efficient and accurate ascertainment of copy number variations (CNVs) at the population level is essential to understand the evolutionary process and population genetics, and to apply CNVs in population-based genome-wide association studies for complex human diseases. We propose a novel Bayesian segmentation approach to identify CNVs in a defined population of any size. It is computationally efficient and provides statistical evidence for the detected CNVs through the Bayes factor. This approach has the unique feature of carrying out segmentation and assigning copy number status simultaneously-a desirable property that current segmentation methods do not share. RESULTS: In comparisons with popular two-step segmentation methods for a single individual using benchmark simulation studies, we find the new approach to perform competitively with respect to false discovery rate and sensitivity in breakpoint detection. In a simulation study of multiple samples with recurrent copy numbers, the new approach outperforms two leading single sample methods. We further demonstrate the effectiveness of our approach in population-level analysis of previously published HapMap data. We also apply our approach in studying population genetics of CNVs. AVAILABILITY: R programs are available at http://www.mshri.on.ca/mitacs/software/SOFTWARE.HTML

Locus-specific heritability estimation via the bootstrap in linkage scans for quantitative trait loci.

BACKGROUND/AIMS: In genome-wide linkage analysis of quantitative trait loci (QTL), locus-specific heritability estimates are biased when the original data are used to both localize linkage and estimate effects, due to maximization of the LOD score over the genome. Positive bias is increased by adoption of stringent significance levels to control genome-wide type I error. We propose multi-locus bootstrap resampling estimators for bias reduction in the situation in which linkage peaks at more than one QTL are of interest. METHODS: Bootstrap estimates were based on repeated sample splitting in the original dataset. We conducted simulation studies in nuclear families with 0 to 5 QTLs and applied the methods in a genome-wide analysis of a blood pressure phenotype in extended pedigrees from the Framingham Heart Study (FHS). RESULTS: Compared to naïve estimates in the original simulation samples, bootstrap estimates had reduced bias and smaller mean squared error. In the FHS pedigrees, the bootstrap yielded heritability estimates as much as 70% smaller than in the original sample. CONCLUSIONS: Because effect estimates obtained in an initial study are typically inflated relative to those expected in an independent replication study, successful replication will be more likely when sample size requirements are based on bias-reduced estimates.

Resampling methods to reduce the selection bias in genetic effect estimation in genome-wide scans.

Using the simulated data of Problem 2 for Genetic Analysis Workshop 14 (GAW14), we investigated the ability of three bootstrap-based resampling estimators (a shrinkage, an out-of-sample, and a weighted estimator) to reduce the selection bias for genetic effect estimation in genome-wide linkage scans. For the given marker density in the preliminary genome scans (7 cM for microsatellite and 3 cM for SNP), we found that the two sets of markers produce comparable results in terms of power to detect linkage, localization accuracy, and magnitude of test statistic at the peak location. At the locations detected in the scan, application of the three bootstrap-based estimators substantially reduced the upward selection bias in genetic effect estimation for both true and false positives. The relative effectiveness of the estimators depended on the true genetic effect size and the inherent power to detect it. The shrinkage estimator is recommended when the power to detect the disease locus is low. Otherwise, the weighted estimator is recommended.

Comparison of family-based association tests in chromosome regions selected by linkage-based confidence intervals.

We use the Genetic Analysis Workshop 14 simulated data to explore the effectiveness of a two-stage strategy for mapping complex disease loci consisting of an initial genome scan with confidence interval construction for gene location, followed by fine mapping with family-based tests of association on a dense set of single-nucleotide polymorphisms. We considered four types of intervals: the 1-LOD interval, a basic percentile bootstrap confidence interval based on the position of the maximum Zlr score, and asymptotic and bootstrap confidence intervals based on a generalized estimating equations method. For fine mapping we considered two family-based tests of association: a test based on a likelihood ratio statistic and a transmission-disequilibrium-type test implemented in the software FBAT. In two of the simulation replicates, we found that the bootstrap confidence intervals based on the peak Zlr and the 1-LOD support interval always contained the true disease loci and that the likelihood ratio test provided further strong confirmatory evidence of the presence of disease loci in these regions.