Multivariate Phenotypes Research Articles

Person-Based Brain Morphometric Similarity is Heritable and Correlates With Biological Features

The characterization of the functional significance of interindividual variation in brain morphometry is a core aim of cognitive neuroscience. Prior research has focused on interindividual variation at the level of regional brain measures thus overlooking the fact that each individual brain is a person-specific ensemble of interdependent regions. To expand this line of inquiry we introduce the person-based similarity index (PBSI) for brain morphometry. The conceptual unit of the PBSI is the individual person’s brain structural profile which considers all relevant morphometric measures as features of a single vector. In 2 independent cohorts (total of 1756 healthy participants), we demonstrate the foundational validity of this approach by affirming that the PBSI scores for subcortical volume and cortical thickness in healthy individuals differ between men and women, are heritable, and robust to variation in neuroimaging parameters, sample composition, and regional brain morphometry. Moreover, the PBSI scores correlate with age, body mass index, and fluid intelligence. Collectively, these results suggest that the person-based measures of brain morphometry are biologically and functionally meaningful and have the potential to advance the study of human variation in multivariate brain imaging phenotypes in healthy and clinical populations.

Multivariate analysis of genome-wide data to identify potential pleiotropic genes for type 2 diabetes, obesity and coronary artery disease using MetaCCA

BackgroundAlthough genome-wide association studies (GWAS) have been extensively applied in identifying SNP associated with metabolic diseases, the SNPs identified by this prevailing univariate approach only explain a small percentage of the genetic variance of traits. The extensive previous studies have repeatedly shown type2 diabetes (T2D), obesity and coronary artery disease (CAD) have common genetic mechanisms and the overlapping pathophysiological pathways. MethodsThe genetic pleiotropy-informed metaCCA method was applied on summary statistics data from three independent meta-GWAS summary statistics to identify shared variants and pleiotropic effect between T2D, obesity and CAD. Furthermore, to refine all genes, we performed gene-based association analyses for these three diseases respectively using VEGAS2. Gene enrichment analysis was applied to explore the potential functional significance of the identified genes. ResultsAfter metaCCA analysis, 833 SNPs reached the Bonferroni corrected threshold (p < 7.99 × 10−7) in the univariate SNP-multivariate phenotype analysis, and 327 genes with a significance threshold (p < 3.73 × 10−6) were identified as potentially pleiotropic genes in the multivariate SNP-multivariate phenotype analysis. By screening the results of gene-based p-values, we identified 22 putative pleiotropic genes which achieved significance threshold in metaCCA analyses and were also associated with at least one disease in the VEGAS2 analyses. ConclusionsThe metaCCA method identified novel variants associated with T2D, obesity and CAD by effectively incorporating information from different GWAS datasets. Our analyses may provide insights for some common therapeutic approaches of metabolic diseases based on the pleiotropic genes and common mechanisms identified.

Experimental manipulation of a signal trait reveals complex phenotype-behaviour coordination

Animals use morphological signals such as ornamental traits or weaponry to mediate social interactions, and the extent of signal trait elaboration is often positively associated with reproductive success. By demonstrating relationships between signal traits and fitness, researchers often make inferences about how behaviour operates to shape those outcomes. However, detailed information about fine-scale individual behaviour, and its physiological basis, can be difficult to obtain. Here we show that experimental manipulations to exaggerate a signal trait (plumage colour) and concomitant changes in testosterone and stress-induced corticosterone levels altered social interactivity between manipulated males and their social mates. On average, darkened males did not have higher levels of interactivity than unmanipulated males; however, males who experienced a greater shift in colour (pale to dark), a larger, positive change in testosterone levels, and a dampened stress-induced corticosterone response had a larger increase in the number of interactions with their social mate post-manipulation compared to pre-manipulation. This work provides new insights into the integration and real-time flexibility of multivariate phenotypes and direct evidence for the role of social interactions in pair bond maintenance.

Development of G: a test in an amphibious fish.

Heritable variation in, and genetic correlations among, traits determine the response of multivariate phenotypes to natural selection. However, as traits develop over ontogeny, patterns of genetic (co)variation and integration captured by the G matrix may also change. Despite this, few studies have investigated how genetic parameters underpinning multivariate phenotypes change as animals pass through major life history stages. Here, using a self-fertilizing hermaphroditic fish species, mangrove rivulus (Kryptolebias marmoratus), we test the hypothesis that G changes from hatching through reproductive maturation. We also test Cheverud's conjecture by asking whether phenotypic patterns provide an acceptable surrogate for patterns of genetic (co)variation within and across ontogenetic stages. For a set of morphological traits linked to locomotor (jumping) performance, we find that the overall level of genetic integration (as measured by the mean-squared correlation across all traits) does not change significantly over ontogeny. However, we also find evidence that some trait-specific genetic variances and pairwise genetic correlations do change. Ontogenetic changes in G indicate the presence of genetic variance for developmental processes themselves, while also suggesting that any genetic constraints on morphological evolution may be age-dependent. Phenotypic correlations closely resembled genetic correlations at each stage in ontogeny. Thus, our results are consistent with the premise that-at least under common environment conditions-phenotypic correlations can be a good substitute for genetic correlations in studies of multivariate developmental evolution.

Comparisons of gut microbiota profiles in wild-type and gelatinase B/matrix metalloproteinase-9-deficient mice in acute DSS-induced colitis

Gut microbiota help to educate the immune system and a number of involved immune cells were recently characterized. However, specific molecular determinants in these processes are not known, and, reciprocally, little information exists about single host determinants that alter the microbiota. Gelatinase B/matrix metalloproteinase-9 (MMP-9), an innate immune regulator and effector, has been suggested as such a host determinant. In this study, acute colitis was induced in co-housed MMP-9-/- mice (n = 10) and their wild-type (WT) littermates (n = 10) via oral administration of 3% dextran sodium sulfate (DSS) for 7 days followed by 2 days of regular drinking water. Control mice (10 WT and 10 MMP-9-/-) received normal drinking water. Fecal samples were collected at time of sacrifice and immediately frozen at −80 °C. Microbiota analysis was performed using 16S rRNA amplicon sequencing on Illumina MiSeq and taxonomic annotation was performed using the Ribosomal Database Project as reference. Statistical analysis correcting for multiple testing was done using R. No significant differences in clinical or histopathological parameters were found between both genotypes with DSS-induced colitis. Observed microbial richness at genus level and microbiota composition were not significantly influenced by host genotype. In contrast, weight loss, disease activity index, cage, and phenotype did significantly influence the intestinal microbiota composition. After multivariate analysis, cage and phenotype were identified as the sole drivers of microbiota composition variability. In conclusion, changes in fecal microbiota composition were not significantly altered in MMP-9-deficient mice compared to wild-type littermates, but instead were mainly driven by DSS-induced colonic inflammation.

Similarities and differences in intrapopulation trait correlations of co-occurring tree species: consistent water-use relationships amid widely different correlation patterns.

General relationships among functional traits have been identified across species, but the forces shaping these relationships remain largely unknown. Adopting an approach from evolutionary biology, we studied similarities and differences in intrapopulation trait correlations among locally co-occurring tree species to assess the roles of constraints, phylogeny, and the environmental niche in shaping multivariate phenotypes. We tested the hypotheses (1) that intrapopulation correlations among functional traits are largely shaped by fundamental trade-offs or constraints and (2) that differences among species reflect adaptation to their environmental niches. We compared pairwise correlations and correlation matrices of 17 key functional traits within and among temperate tree species. These traits describe three well-established trade-off dimensions characterizing interspecific relationships among physiological functions: resource acquisition and conservation; sap transport and mechanical support; and branch architecture. Six trait pairs are consistently correlated within populations. Of these, only one involves dimensionally independent traits: LMA-δ13 C. For all other traits, intrapopulation functional trait correlations are weak, are species-specific, and differ from interspecific correlations. Species intrapopulation correlation matrices are related to neither phylogeny nor environmental niche. The results (1) suggest that the functional design of these species is centered on efficient water use, (2) highlight flexibility in plant functional design across species, and (3) suggest that intrapopulation, local interspecific, and global interspecific correlations are shaped by processes acting at each of these scales.

Multitrait genome association analysis identifies new susceptibility genes for human anthropometric variation in the GCAT cohort

BackgroundHeritability estimates have revealed an important contribution of SNP variants for most common traits; however, SNP analysis by single-trait genome-wide association studies (GWAS) has failed to uncover their impact. In...

Evolution of steroid hormones in reproductive females of the threespine stickleback fish

Hormones play a prominent role in animal development, mediating the expression of traits and coordinating phenotypic responses to the environment. Their role as physiological integrators has implications for how populations respond to natural selection and can impact the speed and direction of evolutionary change. However, many emerging and established fish models with the potential to be ecologically or evolutionarily informative are small-bodied, making hormone sampling through traditional methods (whole-body or plasma) lethal or highly disruptive. Sampling methodology has thus restricted study design, often limiting sample sizes, and has prevented the study of at-risk/endangered populations. We utilize water-borne hormone sampling, a minimally invasive method of measuring the rate of steroid hormone release across the gills and further validate this method in a novel, evolutionary context. First, we compare water-borne hormone measures of cortisol with those quantified from plasma and whole-body samples collected from the same individuals to establish the relationship between concentrations quantified via the three methods. We then compare the release of steroid hormones in three populations of threespine stickleback to establish the sensitivity of this tool in measuring within-individual and between-individual variation in biologically relevant contexts (reproductive stages), and in assessing differences among populations with distinct evolutionary histories. We demonstrate a strong positive relationship between cortisol concentrations measured with water-borne, plasma, and whole-body collection techniques. Tracking estradiol and testosterone throughout clutch production in females produced anticipated patterns associated with growing and maturing eggs, with divergence in estradiol production in one population. Additionally, differences among populations in cortisol levels at ovulation paralleled the relative presence of a social stressor, and thus expected energetic needs within each population. We confirm that water-borne hormone sampling is sufficiently sensitive to capture biologically relevant fluctuations in steroid hormones between environmental contexts and demonstrate that among-population differences are detectable. This technique can be applied broadly to small fish to answer important ecological and evolutionary questions. By linking population variation in hormones and the multivariate phenotype, this technique will help elucidate both proximate mechanisms underlying phenotypic development and variation, and the way hormone networks alter evolutionary responses to selection.

Fatty acids in an iteroparous fish: variable complexity, identity, and phenotypic correlates

Ecologists often focus on summarized composition when assessing complex, multivariate phenotypes such as fatty acids. Increasing complexity in fatty acid composition may offer benefits to individuals that may not be recognized by assessing mean fatty acid identity. We quantified fatty acid identity and complexity in the egg and muscle of spawning female yellow perch (Perca flavescens (Mitchill, 1814)) exposed to three overwinter thermal regimes (4, 8, and 13 °C) to evaluate (i) thermal regulation of fatty acid complexity, (ii) fatty acid complexity and identity differences between female muscle and eggs, and (iii) relationships between fatty acid complexity and egg traits. Temperature did not alter fatty acid complexity in either tissue. Muscle contained an even distribution of 12 abundant polyunsaturated fatty acids (PUFA), while eggs exhibited a simpler phenotype of eight compounds. Fatty acid complexity in female muscle was positively correlated with potential indicators of egg quality, including size and lipid content. Preferential allocation of fatty acids to eggs by females likely contributed to a simple phenotype dominated by PUFA precursors and energy sources, while muscle represented a more complex fatty acid phenotype. Recognizing complexity could elucidate variation in reproductive condition among individuals, while identifying patterns in fatty acid allocation among populations.

Multivariate analysis of genome-wide data to identify potential pleiotropic genes for five major psychiatric disorders using MetaCCA

Background: Genome-wide association studies have been extensively applied in identifying SNP associated with major psychiatric disorders. However, the SNPs identified by the prevailing univariate approach only explain a small percentage of the genetic variance of traits, and the extensive data have shown the major psychiatric disorders have common biological mechanisms and the overlapping pathophysiological pathways. Methods: We applied the genetic pleiotropy-informed metaCCA method on summary statistics data from the Psychiatric Genomics Consortium Cross-Disorder Group to examine the overlapping genetic relations between the five major psychiatric disorders. Furthermore, to refine all genes, we performed gene-based association analyses for the five disorders respectively using VEGAS2. Gene enrichment analysis was applied to explore the potential functional significance of the identified genes. Results: After metaCCA analysis, 1147 SNPs reached the Bonferroni corrected threshold (p < 1.06 × 10−6) in the univariate SNP-multivariate phenotype analysis, and 246 genes with a significance threshold (p < 3.85 × 10−6) were identified as potentially pleiotropic genes in the multivariate SNP-multivariate phenotype analysis. By screening the results of gene-based p-values, we identified 37 putative pleiotropic genes which achieved significance threshold in metaCCA analyses and were also associated with at least one disorder in the VEGAS2 analyses. Limitations: Alternative approaches and experimental studies may be applied to check whether novel genes could still be identified/substantiated with these methods. Conclusions: The metaCCA method identified novel variants associated with psychiatric disorders by effectively incorporating information from different GWAS datasets. Our analyses may provide insights for some common therapeutic approaches of these five major psychiatric disorders based on the pleiotropic genes and common mechanisms identified.

Phylogenetic ANOVA: Group-clade aggregation, biological challenges, and a refined permutation procedure.

Phylogenetic regression is frequently used in macroevolutionary studies, and its statistical properties have been thoroughly investigated. By contrast, phylogenetic ANOVA has received relatively less attention, and the conditions leading to incorrect statistical and biological inferences when comparing multivariate phenotypes among groups remain underexplored. Here, we propose a refined method of randomizing residuals in a permutation procedure (RRPP) for evaluating phenotypic differences among groups while conditioning the data on the phylogeny. We show that RRPP displays appropriate statistical properties for both phylogenetic ANOVA and regression models, and for univariate and multivariate datasets. For ANOVA, we find that RRPP exhibits higher statistical power than methods utilizing phylogenetic simulation. Additionally, we investigate how group dispersion across the phylogeny affects inferences, and reveal that highly aggregated groups generate strong and significant correlations with the phylogeny, which reduce statistical power and subsequently affect biological interpretations. We discuss the broader implications of this phylogenetic group aggregation, and its relation to challenges encountered with other comparative methods where one or a few transitions in discrete traits are observed on the phylogeny. Finally, we recommend that phylogenetic comparative studies of continuous trait data use RRPP for assessing the significance of indicator variables as sources of trait variation.

Whole genome association study of brain-wide imaging phenotypes: A study of the ping cohort.

Neuropsychological disorders have a biological basis rooted in brain function, and neuroimaging data are expected to better illuminate the complex genetic basis of neuropsychological disorders. Because they are biological measures, neuroimaging data avoid biases arising from clinical diagnostic criteria that are subject to human understanding and interpretation. A challenge with analyzing neuroimaging data is their high dimensionality and complex spatial relationships. To tackle this challenge, we introduced a novel distance covariance tests that can assess the association between genetic markers and multivariate diffusion tensor imaging measurements, and analyzed a genome-wide association study (GWAS) dataset collected by the Pediatric Imaging, Neurocognition, and Genetics (PING) study. We also considered existing approaches as comparisons. Our results showed that, after correcting for multiplicity, distance covariance tests of the multivariate phenotype yield significantly greater power at detecting genetic markers affecting brain structure than standard mass univariate GWAS of individual neuroimaging biomarkers. Our results underscore the usefulness of utilizing the distance covariance to incorporate neuroimaging data in GWAS.

Evolution of Genetic Variance during Adaptive Radiation.

Genetic correlations between traits can concentrate genetic variance into fewer phenotypic dimensions that can bias evolutionary trajectories along the axis of greatest genetic variance and away from optimal phenotypes, constraining the rate of evolution. If genetic correlations limit adaptation, rapid adaptive divergence between multiple contrasting environments may be difficult. However, if natural selection increases the frequency of rare alleles after colonization of new environments, an increase in genetic variance in the direction of selection can accelerate adaptive divergence. Here, we explored adaptive divergence of an Australian native wildflower by examining the alignment between divergence in phenotype mean and divergence in genetic variance among four contrasting ecotypes. We found divergence in mean multivariate phenotype along two major axes represented by different combinations of plant architecture and leaf traits. Ecotypes also showed divergence in the level of genetic variance in individual traits and the multivariate distribution of genetic variance among traits. Divergence in multivariate phenotypic mean aligned with divergence in genetic variance, with much of the divergence in phenotype among ecotypes associated with changes in trait combinations containing substantial levels of genetic variance. Overall, our results suggest that natural selection can alter the distribution of genetic variance underlying phenotypic traits, increasing the amount of genetic variance in the direction of natural selection and potentially facilitating rapid adaptive divergence during an adaptive radiation.

Exome-wide rare variant analyses of two bone mineral density phenotypes: the challenges of analyzing rare genetic variation

Performance of a recently developed test for association between multivariate phenotypes and sets of genetic variants (MURAT) is demonstrated using measures of bone mineral density (BMD). By combining individual-level whole genome sequenced data from the UK10K study, and imputed genome-wide genetic data on individuals from the Study of Osteoporotic Fractures (SOF) and the Osteoporotic Fractures in Men Study (MrOS), a data set of 8810 individuals was assembled; tests of association were performed between autosomal gene-sets of genetic variants and BMD measured at lumbar spine and femoral neck. Distributions of p-values obtained from analyses of a single BMD phenotype are compared to those from the multivariate tests, across several region definitions and variant weightings. There is evidence of increased power with the multivariate test, although no new loci for BMD were identified. Among 17 genes highlighted either because there were significant p-values in region-based association tests or because they were in well-known BMD genes, 4 windows in 2 genes as well as 6 single SNPs in one of these genes showed association at genome-wide significant thresholds with the multivariate phenotype test but not with the single-phenotype test, Sequence Kernel Association Test (SKAT).

Evolutionary dynamics of the leaf phenological cycle in an oak metapopulation along an elevation gradient.

It has been predicted that environmental changes will radically alter the selective pressures on phenological traits. Long-lived species, such as trees, will be particularly affected, as they may need to undergo major adaptive change over only one or a few generations. The traits describing the annual life cycle of trees are generally highly evolvable, but nothing is known about the strength of their genetic correlations. Tight correlations can impose strong evolutionary constraints, potentially hampering the adaptation of multivariate phenological phenotypes. In this study, we investigated the evolutionary, genetic and environmental components of the timing of leaf unfolding and senescence within an oak metapopulation along an elevation gradient. Population divergence, estimated from insitu and common-garden data, was compared to expectations under neutral evolution, based on microsatellite markers. This approach made it possible (1) to evaluate the influence of genetic correlation on multivariate local adaptation to elevation and (2) to identify traits probably exposed to past selective pressures due to the colder climate at high elevation. The genetic correlation was positive but very weak, indicating that genetic constraints did not shape the local adaptation pattern for leaf phenology. Both spring and fall (leaf unfolding and senescence, respectively) phenology timings were involved in local adaptation, but leaf unfolding was probably the trait most exposed to climate change-induced selection. Our data indicated that genetic variation makes a much smaller contribution to adaptation than the considerable plastic variation displayed by a tree during its lifetime. The evolutionary potential of leaf phenology is, therefore, probably not the most critical aspect for short-term population survival in a changing climate.

Identifying pleiotropic genes in genome-wide association studies from related subjects using the linear mixed model and Fisher combination function

BackgroundA multivariate genome-wide association test is proposed for analyzing data on multivariate quantitative phenotypes collected from related subjects. The proposed method is a two-step approach. The first step models the association between the genotype and marginal phenotype using a linear mixed model. The second step uses the correlation between residuals of the linear mixed model to estimate the null distribution of the Fisher combination test statistic.ResultsThe simulation results show that the proposed method controls the type I error rate and is more powerful than the marginal tests across different population structures (admixed or non-admixed) and relatedness (related or independent). The statistical analysis on the database of the Study of Addiction: Genetics and Environment (SAGE) demonstrates that applying the multivariate association test may facilitate identification of the pleiotropic genes contributing to the risk for alcohol dependence commonly expressed by four correlated phenotypes.ConclusionsThis study proposes a multivariate method for identifying pleiotropic genes while adjusting for cryptic relatedness and population structure between subjects. The two-step approach is not only powerful but also computationally efficient even when the number of subjects and the number of phenotypes are both very large.

Accounting for Sampling Error in Genetic Eigenvalues Using Random Matrix Theory.

The distribution of genetic variance in multivariate phenotypes is characterized by the empirical spectral distribution of the eigenvalues of the genetic covariance matrix. Empirical estimates of genetic eigenvalues from random effects linear models are known to be overdispersed by sampling error, where large eigenvalues are biased upward, and small eigenvalues are biased downward. The overdispersion of the leading eigenvalues of sample covariance matrices have been demonstrated to conform to the Tracy-Widom (TW) distribution. Here we show that genetic eigenvalues estimated using restricted maximum likelihood (REML) in a multivariate random effects model with an unconstrained genetic covariance structure will also conform to the TW distribution after empirical scaling and centering. However, where estimation procedures using either REML or MCMC impose boundary constraints, the resulting genetic eigenvalues tend not be TW distributed. We show how using confidence intervals from sampling distributions of genetic eigenvalues without reference to the TW distribution is insufficient protection against mistaking sampling error as genetic variance, particularly when eigenvalues are small. By scaling such sampling distributions to the appropriate TW distribution, the critical value of the TW statistic can be used to determine if the magnitude of a genetic eigenvalue exceeds the sampling error for each eigenvalue in the spectral distribution of a given genetic covariance matrix.

Genome-wide association study of facial morphology reveals novel associations with FREM1 and PARK2.

Several studies have now shown evidence of association between common genetic variants and quantitative facial traits in humans. The reported associations generally involve simple univariate measures and likely represent only a small fraction of the genetic loci influencing facial morphology. In this study, we applied factor analysis to a set of 276 facial linear distances derived from 3D facial surface images of 2187 unrelated individuals of European ancestry. We retained 23 facial factors, which we then tested for genetic associations using a genome-wide panel of 10,677,593 single nucleotide polymorphisms (SNPs). In total, we identified genome-wide significant (p < 5 × 10−8) associations in three regions, including two that are novel: one involving measures of midface height at 6q26 within an intron of PARK2 (lead SNP rs9456748; p = 4.99 × 10−8) and another involving measures of central upper lip height at 9p22 within FREM1 (lead SNP rs72713618; p = 2.02 × 10−8). In both cases, the genetic association was stronger with the composite facial factor phenotype than with any of the individual linear distances that comprise those factors. While the biological role of PARK2 in the craniofacial complex is currently unclear, there is evidence from both mouse models and Mendelian syndromes that FREM1 may influence facial variation. These results highlight the potential value of data-driven multivariate phenotyping for genetic studies of human facial morphology.

Bayesian longitudinal low-rank regression models for imaging genetic data from longitudinal studies

To perform a joint analysis of multivariate neuroimaging phenotypes and candidate genetic markers obtained from longitudinal studies, we develop a Bayesian longitudinal low-rank regression (L2R2) model. The L2R2 model integrates three key methodologies: a low-rank matrix for approximating the high-dimensional regression coefficient matrices corresponding to the genetic main effects and their interactions with time, penalized splines for characterizing the overall time effect, and a sparse factor analysis model coupled with random effects for capturing within-subject spatio-temporal correlations of longitudinal phenotypes. Posterior computation proceeds via an efficient Markov chain Monte Carlo algorithm. Simulations show that the L2R2 model outperforms several other competing methods. We apply the L2R2 model to investigate the effect of single nucleotide polymorphisms (SNPs) on the top 10 and top 40 previously reported Alzheimer disease-associated genes. We also identify associations between the interactions of these SNPs with patient age and the tissue volumes of 93 regions of interest from patients’ brain images obtained from the Alzheimer's Disease Neuroimaging Initiative.

Identifying Pleiotropic Genes in Genome-Wide Association Studies for Multivariate Phenotypes with Mixed Measurement Scales.

We propose a multivariate genome-wide association test for mixed continuous, binary, and ordinal phenotypes. A latent response model is used to estimate the correlation between phenotypes with different measurement scales so that the empirical distribution of the Fisher’s combination statistic under the null hypothesis is estimated efficiently. The simulation study shows that our proposed correlation estimation methods have high levels of accuracy. More importantly, our approach conservatively estimates the variance of the test statistic so that the type I error rate is controlled. The simulation also shows that the proposed test maintains the power at the level very close to that of the ideal analysis based on known latent phenotypes while controlling the type I error. In contrast, conventional approaches–dichotomizing all observed phenotypes or treating them as continuous variables–could either reduce the power or employ a linear regression model unfit for the data. Furthermore, the statistical analysis on the database of the Study of Addiction: Genetics and Environment (SAGE) demonstrates that conducting a multivariate test on multiple phenotypes can increase the power of identifying markers that may not be, otherwise, chosen using marginal tests. The proposed method also offers a new approach to analyzing the Fagerström Test for Nicotine Dependence as multivariate phenotypes in genome-wide association studies.