Harald H. H. Goring

Statistical analysis of multinomial data in complex datasets often requires estimation of the multivariate normal (mvn) distribution for models in which the dimensionality can easily reach 10–1000 and higher. Few algorithms for estimating the mvn distribution can offer robust and efficient performance over such a range of dimensions. We report a simulation-based comparison of two algorithms for the mvn that are widely used in statistical genetic applications. The venerable Mendell-Elston approximation is fast but execution time increases rapidly with the number of dimensions, estimates are generally biased, and an error bound is lacking. The correlation between variables significantly affects absolute error but not overall execution time. The Monte Carlo-based approach described by Genz returns unbiased and error-bounded estimates, but execution time is more sensitive to the correlation between variables. For ultra-high-dimensional problems, however, the Genz algorithm exhibits better scale characteristics and greater time-weighted efficiency of estimation.

High-throughput platforms allow the characterization of thousands of previously known methylation sites. These platforms have great potential for investigating the epigenetic effects that are partially responsible for gene expression control. Methylation sites provide a bridge for the investigation of real-time environmental contributions on genomic events by the alteration of methylation status of those sites. Using the data provided by GAW20’s organization committee, we calculated the heritability estimates of each cytosine-phosphate-guanine (CpG) island before and after the use of fenofibrate, a lipid-control drug. Surprisingly, we detected substantially high heritability estimates before drug usage. This somewhat unexpected high sample correlation was corrected by the use of principal components and the distributions of heritability estimates before and after fenofibrate treatment, which made the distributions comparable. The methylation sites located near a gene were collected and a genetic relationship matrix estimated to represent the overall correlation between samples. We implemented a random-effect association test to screen genes whose methylation patterns partially explain the observable high-density lipoprotein (HDL) heritability. Our leading association was observed for the TMEM52 gene that encodes a transmembrane protein, and is largely expressed in the liver, had not been previously associated with HDL until this manuscript. Using a variance component decomposition framework with the linear mixed model allows the integration of data from different sources, such as methylation, gene expression, metabolomics, and proteomics. The decomposition of the genetic variance component decomposition provides a flexible analytical approach for the challenges of this new omics era.

Phasing, the process of predicting haplotypes from genotype data, is an important undertaking in genetics and an ongoing area of research. Phasing methods, and associated software, designed specifically for pedigrees are urgently needed. Here we present a new method for phasing genotypes from whole genome sequencing data in pedigrees: PULSAR ( P hasing U sing L ineage S pecific A lleles / R are variants). The method is built upon the idea that alleles that are specific to a single founding chromosome within a pedigree, which we refer to as lineage-specific alleles, are highly informative for identifying haplotypes that are identical-by-decent between individuals within a pedigree. Through extensive simulation we assess the performance of PULSAR in a variety of pedigree sizes and structures, and we explore the effects of genotyping errors and presence of non-sequenced individuals on its performance. If the genotyping error rate is sufficiently low PULSAR can phase > 99.9% of heterozygous genotypes with a switch error rate below 1 x 10 ^-4 in pedigrees where all individuals are sequenced. We demonstrate that the method is highly accurate and consistently outperforms the long-range phasing approach used for comparison in our benchmarking. The method also holds promise for fixing genotype errors or imputing missing genotypes. The software implementation of this method is freely available.

The Genetic Analysis Workshops (GAW) are a forum for development, testing, and comparison of statistical genetic methods and software. Each contribution to the workshop includes an application to a specified data set. Here we describe the data distributed for GAW19, which focused on analysis of human genomic and transcriptomic data. GAW19 data were donated by the T2D-GENES Consortium and the San Antonio Family Heart Study and included whole genome and exome sequences for odd-numbered autosomes, measures of gene expression, systolic and diastolic blood pressures, and related covariates in two Mexican American samples. These two samples were a collection of 20 large families with whole genome sequence and transcriptomic data and a set of 1943 unrelated individuals with exome sequence. For each sample, simulated phenotypes were constructed based on the real sequence data. ‘Functional’ genes and variants for the simulations were chosen based on observed correlations between gene expression and blood pressure. The simulations focused primarily on additive genetic models but also included a genotype-by-medication interaction. A total of 245 genes were designated as ‘functional’ in the simulations with a few genes of large effect and most genes explaining < 1 % of the trait variation. An additional phenotype, Q1, was simulated to be correlated among related individuals, based on theoretical or empirical kinship matrices, but was not associated with any sequence variants. Two hundred replicates of the phenotypes were simulated. The GAW19 data are an expansion of the data used at GAW18, which included the family-based whole genome sequence, blood pressure, and simulated phenotypes, but not the gene expression data or the set of 1943 unrelated individuals with exome sequence.

The variation in serum uric acid concentrations is under significant genetic influence. Elevated SUA concentrations have been linked to increased risk for gout, kidney stones, chronic kidney disease, and cardiovascular disease whereas reduced serum uric acid concentrations have been linked to multiple sclerosis, Parkinson’s disease and Alzheimer’s disease. Previously, we identified a novel locus on chromosome 3p26 affecting serum uric acid concentrations in Mexican Americans from San Antonio Family Heart Study. As a follow up, we examined genome-wide single nucleotide polymorphism data in an extended cohort of 1281 Mexican Americans from multigenerational families of the San Antonio Family Heart Study and the San Antonio Family Diabetes/Gallbladder Study. We used a linear regression-based joint linkage/association test under an additive model of allelic effect, while accounting for non-independence among family members via a kinship variance component. Univariate genetic analysis indicated serum uric acid concentrations to be significant heritable (h 2 = 0.50 ± 0.05, p < 4 × 10−35), and linkage analysis of serum uric acid concentrations confirmed our previous finding of a novel locus on 3p26 (LOD = 4.9, p < 1 × 10−5) in the extended sample. Additionally, we observed strong association of serum uric acid concentrations with variants in following candidate genes in the 3p26 region; inositol 1,4,5-trisphosphate receptor, type 1 (ITPR1), contactin 4 (CNTN4), decapping mRNA 1A (DCP1A); transglutaminase 4 (TGM4) and rho guanine nucleotide exchange factor (GEF) 26 (ARHGEF26) [p < 3 × 10−7; minor allele frequencies ranged between 0.003 and 0.42] and evidence of cis-regulation for ITPR1 transcripts. Our results confirm the importance of the chromosome 3p26 locus and genetic variants in this region in the regulation of serum uric acid concentrations.

There is growing interest in the hypertriglyceridemic waist (HTGW) phenotype, defined as high waist circumference (≥95 cm in males and ≥80 cm in females) combined with high serum triglyceride concentration (≥2.0 mmol/L in males and ≥1.5 mmol/L in females) as a marker of type 2 diabetes (T2D) and cardiovascular disease. However, the prevalence of this phenotype in high-risk populations, its association with T2D, and the genetic or epigenetic influences on HTGW are not well explored. Using data from large, extended families of Mexican Americans (a high-risk minority population in the USA) we aimed to: (1) estimate the prevalence of this phenotype, (2) test its association with T2D and related traits, and (3) dissect out the genetic and epigenetic associations with this phenotype using genome-wide and epigenome-wide studies, respectively. Data for this study was from 850 Mexican American participants (representing 39 families) recruited under the ongoing San Antonio Family Heart Study, 26 % of these individuals had HTGW. This phenotype was significantly heritable (h 2 r = 0.52, p = 1.1 × 10−5) and independently associated with T2D as well as fasting glucose levels and insulin resistance. We conducted genome-wide association analyses using 759,809 single nucleotide polymorphisms (SNPs) and epigenome-wide association analyses using 457,331 CpG sites. There was no evidence of any SNP associated with HTGW at the genome-wide level but two CpG sites (cg00574958 and cg17058475) in CPT1A and one CpG site (cg06500161) in ABCG1 were significantly associated with HTGW and remained significant after adjusting for the closely related components of metabolic syndrome. CPT1A holds a cardinal position in the metabolism of long-chain fatty acids while ABCG1 plays a role in triglyceride metabolism. Our results reemphasize the value of HTGW as a marker of T2D. This phenotype shows association with DNA methylation within CPT1A and ABCG1, genes involved in fatty acid and triglyceride metabolism. Our results underscore the importance of epigenetics in a clinically informative phenotype.