Three Approaches to Modeling Gene-Environment Interactions in Longitudinal Family Data: Gene-Smoking Interactions in Blood Pressure.

Blood pressure (BP) has been shown to be substantially heritable, yet identified genetic variants explain only a small fraction of the heritability. Gene-smoking interactions have detected novel BP loci in cross-sectional family data. Longitudinal family data are available and have additional promise to identify BP loci. However, this type of data presents unique analysis challenges. Although several methods for analyzing longitudinal family data are available, which method is the most appropriate and under what conditions has not been fully studied. Using data from three clinic visits from the Framingham Heart Study, we performed association analysis accounting for gene-smoking interactions in BP at 31,203 markers on chromosome 22. We evaluated three different modeling frameworks: generalized estimating equations (GEE), hierarchical linear modeling, and pedigree-based mixed modeling. The three models performed somewhat comparably, with multiple overlaps in the most strongly associated loci from each model. Loci with the greatest significance were more strongly supported in the longitudinal analyses than in any of the component single-visit analyses. The pedigree-based mixed model was more conservative, with less inflation in the variant main effect and greater deflation in the gene-smoking interactions. The GEE, but not the other two models, resulted in substantial inflation in the tail of the distribution when variants with minor allele frequency <1% were included in the analysis. The choice of analysis method should depend on the model and the structure and complexity of the familial and longitudinal data.

Influence of Smoking Status and Intensity on Discovery of Blood Pressure Loci Through Gene-Smoking Interactions.

BACKGROUND: Genetic variation accounts for approximately 30% of blood pressure (BP) variability but most of that variability has not been attributed to specific variants. Interactions between genes and BP-associated factors may explain some "missing heritability." Cigarette smoking increases BP after short-term exposure and decreases BP with longer exposure. Gene-smoking interactions have discovered novel BP loci, but the contribution of smoking status and intensity to gene discovery is unknown. METHODS: We analyzed gene-smoking intensity interactions for association with systolic BP (SBP) in three subgroups from the Framingham Heart Study: current smokers only (N = 1,057), current and former smokers ("ever smokers," N = 3,374), and all subjects (N = 6,710). We used three smoking intensity variables defined at cutoffs of 10, 15, and 20 cigarettes per day (CPD). We evaluated the 1 degree-of-freedom (df) interaction and 2df joint test using generalized estimating equations. RESULTS: Analysis of current smokers using a CPD cutoff of 10 produced two loci associated with SBP. The rs9399633 minor allele was associated with increased SBP (5 mmHg) in heavy smokers (CPD > 10) but decreased SBP (7 mmHg) in light smokers (CPD ≤ 10). The rs11717948 minor allele was associated with decreased SBP (8 mmHg) in light smokers but decreased SBP (2 mmHg) in heavy smokers. Across all nine analyses, 19 additional loci reached P < 1 × 10(-6). DISCUSSION: Analysis of current smokers may have the highest power to detect gene-smoking interactions, despite the reduced sample size. Associations of loci near SASH1 and KLHL6/KLHL24 with SBP may be modulated by tobacco smoking.

The role of rare variants in systolic blood pressure: analysis of ExomeChip data in HyperGEN African Americans.

Cardiovascular diseases are among the most significant health problems in the United States today, with their major risk factor, hypertension, disproportionately affecting African Americans (AAs). Although GWAS have identified dozens of common variants associated with blood pressure (BP) and hypertension in European Americans, these variants collectively explain <2.5% of BP variance, and most of the genetic variants remain yet to be identified. Here, we report the results from rare-variant analysis of systolic BP using 94,595 rare and low-frequency variants (minor allele frequency, MAF, <5%) from the Illumina exome array genotyped in 2,045 HyperGEN AAs. In addition to single-variant analysis, 4 gene-level association tests were used for analysis: burden and family-based SKAT tests using MAF cutoffs of 1 and 5%. The gene-based methods often provided lower p values than the single-variant approach. Some consistency was observed across these 4 gene-based analysis options. While neither the gene-based analyses nor the single-variant analysis produced genome-wide significant results, the top signals, which had supporting evidence from multiple gene-based methods, were of borderline significance. Though additional molecular validations are required, 6 of the 16 most promising genes are biologically plausible with physiological connections to BP regulation.

Between candidate genes and whole genomes: time for alternative approaches in blood pressure genetics.

Blood pressure has a significant genetic component, but less than 3% of the observed variance has been attributed to genetic variants identified to date. Candidate gene studies of rare, monogenic hypertensive syndromes have conclusively implicated several genes altering renal sodium balance, and studies of essential hypertension have inconsistently implicated over 50 genes in pathways affecting renal sodium balance and other functions. Genome-wide linkage scans have replicated numerous quantitative trait loci throughout the genome, and over 50 single nucleotide polymorphisms (SNPs) have been replicated in multiple genome-wide association studies. These studies provide considerable evidence that epistasis and other interactions play a role in the genetic architecture of blood pressure regulation, but candidate gene studies have limited scope to test for epistasis, and genome-wide studies have low power for both main effects and interactions. This review summarizes the genetic findings to date for blood pressure, and it proposes focused, pathway-based approaches involving epistasis, gene-environment interactions, and next-generation sequencing to further the genetic dissection of blood pressure and hypertension.

Decoupling of the minority PhD talent pool and assistant professor hiring in medical school basic science departments in the US.

Faculty diversity is a longstanding challenge in the US. However, we lack a quantitative and systemic understanding of how the career transitions into assistant professor positions of PhD scientists from underrepresented minority (URM) and well-represented (WR) racial/ethnic backgrounds compare. Between 1980 and 2013, the number of PhD graduates from URM backgrounds increased by a factor of 9.3, compared with a 2.6-fold increase in the number of PhD graduates from WR groups. However, the number of scientists from URM backgrounds hired as assistant professors in medical school basic science departments was not related to the number of potential candidates (R2=0.12, p>0.07), whereas there was a strong correlation between these two numbers for scientists from WR backgrounds (R2=0.48, p<0.0001). We built and validated a conceptual system dynamics model based on these data that explained 79% of the variance in the hiring of assistant professors and posited no hiring discrimination. Simulations show that, given current transition rates of scientists from URM backgrounds to faculty positions, faculty diversity would not increase significantly through the year 2080 even in the context of an exponential growth in the population of PhD graduates from URM backgrounds, or significant increases in the number of faculty positions. Instead, the simulations showed that diversity increased as more postdoctoral candidates from URM backgrounds transitioned onto the market and were hired.

Single nucleotide polymorphism-single nucleotide polymorphism interactions among inflammation genes in the genetic architecture of blood pressure in the Framingham Heart Study.

BACKGROUND: Hypertension is a major global health burden, but, although systolic and diastolic blood pressure (BP) each have estimated heritability of at least 30%, <3% of their variance has been attributed to particular genetic variants. Few studies have shown interactions between pairs of single nucleotide polymorphisms (SNPs) to be associated with BP. Although many studies use a Bonferroni correction for multiple testing to control type I error, thereby potentially reducing power, false discovery rate (FDR) approaches are also used in genome-wide studies. Renal ion balance genes have been associated with BP regulation, but, although inflammation has been studied in connection with BP, few studies have reported associations between inflammation genes and BP. METHODS: We analyzed SNP-SNP interactions among 31 SNPs from genes involved in renal ion balance and 30 SNPs from genes involved in inflammation using data from the Framingham Heart Study. RESULTS: No evidence of association was found for interactions among renal ion balance SNPs for either systolic or diastolic BP. A group of 3 interactions involving 6 inflammation genes (IKBKB-NFKBIA, IKBKE-CHUK, and ADIPOR2-RETN) showed evidence of association with diastolic BP with an FDR of 4.2%; no single interaction reached experiment-wide significance. CONCLUSIONS: This study identified promising and biologically plausible candidates for interactions between inflammation genes that may be associated with DBP. Analysis using the FDR may allow detection of signals in the presence of modest noise (false positives) that a stringent approach based on Bonferroni-corrected P value thresholds may miss.

Whole genome sequence analysis of the simulated systolic blood pressure in Genetic Analysis Workshop 18 family data: long-term average and collapsing methods.

Analysis of longitudinal family data is challenging because of 2 sources of correlations: correlations across longitudinal measurements and correlations among related individuals. We investigated whether analysis using long-term average (average of all 3 visits) can enhance gene discovery compared with a single-visit analysis. We analyzed all 200 replicates of simulated systolic blood pressure (SBP) in Genetic Analysis Workshop 18 (GAW18) family data using both single-marker and collapsing methods. We considered 2 collapsing approaches: collapsing all variants and collapsing low-frequency variants. Analysis using long-term average performed slightly better than SBP measured at a single visit. Collapsing all variants performed much better than collapsing low-frequency variants at MAP4 and FLNB, which included a common variant with a relatively large effect. For several variants in gene MAP4, single-marker analysis also provided high power. In contrast, collapsing only low-frequency variants performed much better for SCAP, DNASE1L3, and LOC152217, where rare variants in these genes had larger effect than common variants. However, for other causal variants, all approaches provided disappointingly poor performance. This poor performance appeared to occur because most of these causal variants explained a very small fraction of phenotypic variance. We also found that collapsing multiple variants did worse than single-marker analysis for several genes when they contained causal single-nucleotide polymorphisms (SNPs) with both positive and negative effects. Because half of causal SNPs were not found in the annotation file based on the 1000 Genomes Project, we found that power was also affected by our use of incomplete annotation information.

Comparison of two methods for analysis of gene-environment interactions in longitudinal family data: the Framingham heart study.

Gene-environment interaction (GEI) analysis can potentially enhance gene discovery for common complex traits. However, genome-wide interaction analysis is computationally intensive. Moreover, analysis of longitudinal data in families is much more challenging due to the two sources of correlations arising from longitudinal measurements and family relationships. GWIS of longitudinal family data can be a computational bottleneck. Therefore, we compared two methods for analysis of longitudinal family data: a methodologically sound but computationally demanding method using the Kronecker model (KRC) and a computationally more forgiving method using the hierarchical linear model (HLM). The KRC model uses a Kronecker product of an unstructured matrix for correlations among repeated measures (longitudinal) and a compound symmetry matrix for correlations within families at a given visit. The HLM uses an autoregressive covariance matrix for correlations among repeated measures and a random intercept for familial correlations. We compared the two methods using the longitudinal Framingham heart study (FHS) SHARe data. Specifically, we evaluated SNP-alcohol (amount of alcohol consumption) interaction effects on high density lipoprotein cholesterol (HDLC). Keeping the prohibitive computational burden of KRC in mind, we limited the analysis to chromosome 16, where preliminary cross-sectional analysis yielded some interesting results. Our first important finding was that the HLM provided very comparable results but was remarkably faster than the KRC, making HLM the method of choice. Our second finding was that longitudinal analysis provided smaller P-values, thus leading to more significant results, than cross-sectional analysis. This was particularly pronounced in identifying GEIs. We conclude that longitudinal analysis of GEIs is more powerful and that the HLM method is an optimal method of choice as compared to the computationally (prohibitively) intensive KRC method.

Gene-education interactions identify novel blood pressure loci in the Framingham Heart Study.

BACKGROUND: Blood pressure (BP) variability has a genetic component, most of which has yet to be attributed to specific variants. One promising strategy for gene discovery is analysis of interactions between single-nucleotide polymorphisms (SNPs) and BP-related factors, including age, sex, and body mass index (BMI). Educational attainment, a marker for socioeconomic status, has effects on both BP and BMI. METHODS: We investigated SNP-education interaction effects on BP in genome-wide data on 3,836 subjects in families from the Framingham Heart Study. The ABEL suite was used to adjust for age, sex, BMI, medication use, and kinship and to perform 1 degree-of-freedrom (df) and 2 df SNP-education interaction tests. RESULTS: An SNP in PTN was associated with increased systolic BP (5.4mm Hg per minor allele) in those without a bachelor's degree but decreased systolic BP (1.6mm Hg per allele) in those with a bachelor's degree (2 df; P = 2.08 × 10(-8)). An SNP in TOX2 was associated with increased diastolic BP (DBP; 4.1mm Hg per minor allele) in those with no more educational attainment than high school but decreased DBP in those with education past high school (-0.7; 1 df; P = 3.74 × 10(-8)). Three suggestive associations were also found: in MYO16 (pulse pressure: 2 df; P = 2.89 × 10(-7)), in HAS2 (DBP: 1 df; P = 1.41 × 10(-7)), and in DLEU2 (DBP: 2 df; P = 1.93 × 10(-7)). All 5 genes are related to BP, including roles in vasodilation and angiogenesis for PTN and TOX2. CONCLUSIONS: PTN and TOX2 are associated with BP. Analyzing SNP-education interactions may detect novel associations. Education may be a surrogate for unmeasured exposures and behaviors modifying SNP effects on BP.

Multigene interactions and the prediction of depression in the Wisconsin Longitudinal Study.

OBJECTIVES: Single genetic loci offer little predictive power for the identification of depression. This study examined whether an analysis of gene-gene (G × G) interactions of 78 single nucleotide polymorphisms (SNPs) in genes associated with depression and age-related diseases would identify significant interactions with increased predictive power for depression. DESIGN: A retrospective cohort study. SETTING: A survey of participants in the Wisconsin Longitudinal Study. PARTICIPANTS: A total of 4811 persons (2464 women and 2347 men) who provided saliva for genotyping; the group comes from a randomly selected sample of Wisconsin high school graduates from the class of 1957 as well as a randomly selected sibling, almost all of whom are non-Hispanic white. PRIMARY OUTCOME MEASURE: Depression as determine by the Composite International Diagnostic Interview-Short-Form. RESULTS: Using a classification tree approach (recursive partitioning (RP)), the authors identified a number of candidate G × G interactions associated with depression. The primary SNP splits revealed by RP (ANKK1 rs1800497 (also known as DRD2 Taq1A) in men and DRD2 rs224592 in women) were found to be significant as single factors by logistic regression (LR) after controlling for multiple testing (p=0.001 for both). Without considering interaction effects, only one of the five subsequent RP splits reached nominal significance in LR (FTO rs1421085 in women, p=0.008). However, after controlling for G × G interactions by running LR on RP-specific subsets, every split became significant and grew larger in magnitude (OR (before) → (after): men: GNRH1 novel SNP: (1.43 → 1.57); women: APOC3 rs2854116: (1.28 → 1.55), ACVR2B rs3749386: (1.11 → 2.17), FTO rs1421085: (1.32 → 1.65), IL6 rs1800795: (1.12 → 1.85)). CONCLUSIONS: The results suggest that examining G × G interactions improves the identification of genetic associations predictive of depression. 4 of the SNPs identified in these interactions were located in two pathways well known to impact depression: neurotransmitter (ANKK1 and DRD2) and neuroendocrine (GNRH1 and ACVR2B) signalling. This study demonstrates the utility of RP analysis as an efficient and powerful exploratory analysis technique for uncovering genetic and molecular pathway interactions associated with disease aetiology.