Towards fair and clinically relevant polygenic predictions

Lennon et al. recently proposed a clinical polygenic score (PGS) pipeline as part of the Electronic Medical Records and Genomics (eMERGE) network initiative. In this spotlight article we discuss the broader context for the use of PGS in preventive medicine and highlight key limitations and challenges facing their inclusion in prediction models.

Polygenic scores for autism are associated with neurite density in adults and children from the general population

Genetic variants linked to autism are thought to change cognition and behaviour by altering the structure and function of the brain. Although a substantial body of literature has identified structural brain differences in autism, it is unknown whether autism-associated common genetic variants are linked to changes in cortical macro- and micro-structure. We investigated this using neuroimaging and genetic data from adults (UK Biobank, N = 31,748) and children (ABCD, N = 4,928). Using polygenic scores and genetic correlations we observe a robust negative association between common variants for autism and a magnetic resonance imaging derived phenotype for neurite density (intracellular volume fraction) in the general population. This result is consistent across both children and adults, in both the cortex and in white matter tracts, and confirmed using polygenic scores and genetic correlations. There were no sex differences in this association. Mendelian randomisation analyses provide no evidence for a causal relationship between autism and intracellular volume fraction, although this should be revisited using better powered instruments. Overall, this study provides evidence for shared common variant genetics between autism and cortical neurite density.

Contributions of common genetic variants to constitutional delay of puberty and idiopathic hypogonadotropic hypogonadism

CONTEXT

Constitutional delay of puberty (CDP) is highly heritable, but the genetic basis for CDP is largely unknown. Idiopathic hypogonadotropic hypogonadism (IHH) can be caused by rare genetic variants, but in about half of cases, no rare-variant cause is found.

OBJECTIVE

To determine whether common genetic variants that influence pubertal timing contribute to CDP and IHH.

DESIGN

Case-control study.

PARTICIPANTS

80 individuals with CDP; 301 with normosmic IHH, and 348 with Kallmann syndrome; control genotyping data from unrelated studies.

MAIN OUTCOME MEASURES

Polygenic scores (PGS) based on genome-wide association studies for timing of male pubertal hallmarks and age at menarche (AAM).

RESULTS

The CDP cohort had higher PGS for male pubertal hallmarks and for AAM compared to controls (for male hallmarks, Cohen's d = 0.85, p = 1 × 10-16; for AAM, d = 0.67, p = 1 × 10-10). The normosmic IHH cohort also had higher PGS for male hallmarks compared to controls, but the difference was smaller (male hallmarks d = 0.20, p = 0.003; AAM d = 0.10, p = 0.055). No differences were seen for the KS cohort compared to controls (male hallmarks d = 0.04, p = 0.45; AAM d = -0.03, p = 0.86).

CONCLUSIONS

Common genetic variants that influence pubertal timing in the general population contribute strongly to the genetics of CDP, weakly to normosmic IHH, and potentially not at all to KS. These findings demonstrate that the common-variant genetics of CDP and normosmic IHH are largely but not entirely distinct.

Studying the relationship between intelligence quotient and schizophrenia polygenic scores in a family design with first-episode psychosis population

BACKGROUND

The intelligence quotient (IQ) of patients with first-episode psychosis (FEP) and their unaffected relatives may be related to the genetic burden of schizophrenia (SCZ). The polygenic score approach can be useful for testing this question.

AIM

To assess the contribution of the polygenic risk scores for SCZ (PGS-SCZ) and polygenic scores for IQ (PGS-IQ) to the individual IQ and its difference from the mean IQ of the family (named family-IQ) through a family-based design in an FEP sample.

METHODS

The PAFIP-FAMILIES sample (Spain) consists of 122 FEP patients, 131 parents, 94 siblings, and 176 controls. They all completed the WAIS Vocabulary subtest for IQ estimation and provided a DNA sample. We calculated PGS-SCZ and PGS-IQ using the continuous shrinkage method. To account for relatedness in our sample, we performed linear mixed models. We controlled for covariates potentially related to IQ, including age, years of education, sex, and ancestry principal components.

RESULTS

FEP patients significantly deviated from their family-IQ. FEP patients had higher PGS-SCZ than other groups, whereas the relatives had intermediate scores between patients and controls. PGS-IQ did not differ between groups. PGS-SCZ significantly predicted the deviation from family-IQ, whereas PGS-IQ significantly predicted individual IQ.

CONCLUSIONS

PGS-SCZ discriminated between different levels of genetic risk for the disorder and was specifically related to patients' lower IQ in relation to family-IQ. The genetic background of the disorder may affect neurocognition through complex pathological processes interacting with environmental factors that prevent the individual from reaching their familial cognitive potential.

Integrative multi-omics analysis of genomic, epigenomic, and metabolomics data leads to new insights for Attention-Deficit/Hyperactivity Disorder

The evolving field of multi-omics combines data and provides methods for simultaneous analysis across several omics levels. Here, we integrated genomics (transmitted and non-transmitted polygenic scores [PGSs]), epigenomics, and metabolomics data in a multi-omics framework to identify biomarkers for Attention-Deficit/Hyperactivity Disorder (ADHD) and investigated the connections among the three omics levels. We first trained single- and next multi-omics models to differentiate between cases and controls in 596 twins (cases = 14.8%) from the Netherlands Twin Register (NTR) demonstrating reasonable in-sample prediction through cross-validation. The multi-omics model selected 30 PGSs, 143 CpGs, and 90 metabolites. We confirmed previous associations of ADHD with glucocorticoid exposure and the transmembrane protein family TMEM, show that the DNA methylation of the MAD1L1 gene associated with ADHD has a relation with parental smoking behavior, and present novel findings including associations between indirect genetic effects and CpGs of the STAP2 gene. However, out-of-sample prediction in NTR participants (N = 258, cases = 14.3%) and in a clinical sample (N = 145, cases = 51%) did not perform well (range misclassification was [0.40, 0.57]). The results highlighted connections between omics levels, with the strongest connections between non-transmitted PGSs, CpGs, and amino acid levels and show that multi-omics designs considering interrelated omics levels can help unravel the complex biology underlying ADHD.

Using Polygenic Scores for Circadian Rhythms to Predict Wellbeing, Depressive Symptoms, Chronotype, and Health

The association between circadian rhythms and diseases has been well established, while the association with mental health is less explored. Given the heritable nature of circadian rhythms, this study aimed to investigate the relationship between genes underlying circadian rhythms and mental health outcomes, as well as a possible gene-environment correlation for circadian rhythms. Polygenic scores (PGSs) represent the genetic predisposition to develop a certain trait or disease. In a sample from the Netherlands Twin Register (N = 14,021), PGSs were calculated for two circadian rhythm measures: morningness and relative amplitude (RA). The PGSs were used to predict mental health outcomes such as subjective happiness, quality of life, and depressive symptoms. In addition, we performed the same prediction analysis in a within-family design in a subset of dizygotic twins. The PGS for morningness significantly predicted morningness (R2 = 1.55%) and depressive symptoms (R2 = 0.22%). The PGS for RA significantly predicted general health (R2 = 0.12%) and depressive symptoms (R2 = 0.20%). Item analysis of the depressive symptoms showed that 4 out of 14 items were significantly associated with the PGSs. Overall, the results showed that people with a genetic predisposition of being a morning person or with a high RA are likely to have fewer depressive symptoms. The four associated depressive symptoms described symptoms related to decision-making, energy, and feeling worthless or inferior, rather than sleep. Based on our findings future research should include a substantial role for circadian rhythms in depression research and should further explore the gene-environment correlation in circadian rhythms.

Efficacy of federated learning on genomic data: a study on the UK Biobank and the 1000 Genomes Project

Combining training data from multiple sources increases sample size and reduces confounding, leading to more accurate and less biased machine learning models. In healthcare, however, direct pooling of data is often not allowed by data custodians who are accountable for minimizing the exposure of sensitive information. Federated learning offers a promising solution to this problem by training a model in a decentralized manner thus reducing the risks of data leakage. Although there is increasing utilization of federated learning on clinical data, its efficacy on individual-level genomic data has not been studied. This study lays the groundwork for the adoption of federated learning for genomic data by investigating its applicability in two scenarios: phenotype prediction on the UK Biobank data and ancestry prediction on the 1000 Genomes Project data. We show that federated models trained on data split into independent nodes achieve performance close to centralized models, even in the presence of significant inter-node heterogeneity. Additionally, we investigate how federated model accuracy is affected by communication frequency and suggest approaches to reduce computational complexity or communication costs.

Integrative polygenic analysis of the protective effects of fatty acid metabolism on disease as modified by obesity

Dysregulation of fatty acid metabolites can play a crucial role in the progression of complex diseases, such as cardiovascular disease, digestive diseases, and metabolic diseases. Metabolites can have either protective or risk effects on a disease; however, the details of such associations remain contentious. In this study, we demonstrate an integrative PheWAS approach to establish high confidence, causally suggestive of metabolite-disease associations for three fatty acid metabolites, namely, omega-3 fatty acids, omega-6 fatty acids, and docosahexaenoic acid, for 1,254 disease endpoints. Metabolite-disease associations were established if there was a concordant direction of effect and significance for metabolite level and genetic risk score for the metabolite. There was enrichment for metabolite associations with diseases of the respiratory system for omega-3 fatty acids, diseases of the circulatory system and endocrine system for omega-6 fatty acids, and diseases of the digestive system for docosahexaenoic acid. Upon performing Mendelian randomization on a subset of the outcomes, we identified 3, 6, and 15 significant diseases associated with omega-3 fatty acids, omega-6 fatty acids, and docosahexaenoic acid, respectively. We then demonstrate a class of prevalence-risk relationships indicative of (de)canalization of disease under high and low fatty acid metabolite levels. Finally, we show that the interaction between the metabolites and obesity demonstrates that the degree of protection afforded by fatty acid metabolites is strongly modulated by underlying metabolic health. This study evaluated the disease architectures of three polyunsaturated fatty acids (PUFAs), which were validated by several PheWAS modes of support. Our results not only highlight specific diseases associated with each metabolite but also disease group enrichments. In addition, we demonstrate an integrative PheWAS methodology that can be applied to other components of the human metabolome or other traits of interest. The results of this study can be used as an atlas to cross-compare genetic with non-genetic disease associations for the three PUFAs investigated. The findings can be explored through our R shiny app at https://pufa.biosci.gatech.edu.

Genome-wide study investigating effector genes and polygenic prediction for kidney function in persons with ancestry from Africa and the Americas

Chronic kidney disease is a leading cause of death and disability globally and impacts individuals of African ancestry (AFR) or with ancestry in the Americas (AMS) who are under-represented in genome-wide association studies (GWASs) of kidney function. To address this bias, we conducted a large meta-analysis of GWASs of estimated glomerular filtration rate (eGFR) in 145,732 AFR and AMS individuals. We identified 41 loci at genome-wide significance (p < 5 × 10-8), of which two have not been previously reported in any ancestry group. We integrated fine-mapped loci with epigenomic and transcriptomic resources to highlight potential effector genes relevant to kidney physiology and disease, and reveal key regulatory elements and pathways involved in renal function and development. We demonstrate the varying but increased predictive power offered by a multi-ancestry polygenic score for eGFR and highlight the importance of population diversity in GWASs and multi-omics resources to enhance opportunities for clinical translation for all.

Lymphocyte Count Derived Polygenic Score and Interindividual Variability in CD4 T-cell Recovery in Response to Antiretroviral Therapy

Access to safe and effective antiretroviral therapy (ART) is a cornerstone in the global response to the HIV pandemic. Among people living with HIV, there is considerable interindividual variability in absolute CD4 T-cell recovery following initiation of virally suppressive ART. The contribution of host genetics to this variability is not well understood. We explored the contribution of a polygenic score which was derived from large, publicly available summary statistics for absolute lymphocyte count from individuals in the general population (PGSlymph) due to a lack of publicly available summary statistics for CD4 T-cell count. We explored associations with baseline CD4 T-cell count prior to ART initiation (n=4959) and change from baseline to week 48 on ART (n=3274) among treatment-naïve participants in prospective, randomized ART studies of the AIDS Clinical Trials Group. We separately examined an African-ancestry-derived and a European-ancestry-derived PGSlymph, and evaluated their performance across all participants, and also in the African and European ancestral groups separately. Multivariate models that included PGSlymph, baseline plasma HIV-1 RNA, age, sex, and 15 principal components (PCs) of genetic similarity explained ∼26-27% of variability in baseline CD4 T-cell count, but PGSlymph accounted for <1% of this variability. Models that also included baseline CD4 T-cell count explained ∼7-9% of variability in CD4 T-cell count increase on ART, but PGSlymph accounted for <1% of this variability. In univariate analyses, PGSlymph was not significantly associated with baseline or change in CD4 T-cell count. Among individuals of African ancestry, the African PGSlymph term in the multivariate model was significantly associated with change in CD4 T-cell count while not significant in the univariate model. When applied to lymphocyte count in a general medical biobank population (Penn Medicine BioBank), PGSlymph explained ∼6-10% of variability in multivariate models (including age, sex, and PCs) but only ∼1% in univariate models. In summary, a lymphocyte count PGS derived from the general population was not consistently associated with CD4 T-cell recovery on ART. Nonetheless, adjusting for clinical covariates is quite important when estimating such polygenic effects.

Genotype-environment interplay in associations between maternal drinking and offspring emotional and behavioral problems

BACKGROUND

While maternal at-risk drinking is associated with children's emotional and behavioral problems, there is a paucity of research that properly accounts for genetic confounding and gene-environment interplay. Therefore, it remains uncertain what mechanisms underlie these associations. We assess the moderation of associations between maternal at-risk drinking and childhood emotional and behavioral problems by common genetic variants linked to environmental sensitivity (genotype-by-environment [G × E] interaction) while accounting for shared genetic risk between mothers and offspring (GE correlation).

METHODS

We use data from 109 727 children born to 90 873 mothers enrolled in the Norwegian Mother, Father, and Child Cohort Study. Women self-reported alcohol consumption and reported emotional and behavioral problems when children were 1.5/3/5 years old. We included child polygenic scores (PGSs) for traits linked to environmental sensitivity as moderators.

RESULTS

Associations between maternal drinking and child emotional (β1 = 0.04 [95% confidence interval (CI) 0.03-0.05]) and behavioral (β1 = 0.07 [0.06-0.08]) outcomes attenuated after controlling for measured confounders and were almost zero when we accounted for unmeasured confounding (emotional: β1 = 0.01 [0.00-0.02]; behavioral: β1 = 0.01 [0.00-0.02]). We observed no moderation of these adjusted exposure effects by any of the PGS.

CONCLUSIONS

The lack of strong evidence for G × E interaction may indicate that the mechanism is not implicated in this kind of intergenerational association. It may also reflect insufficient power or the relatively benign nature of the exposure in this sample.

Risk and resilience in Syrian refugee children: A multisystem analysis

Refugee children are often exposed to substantial trauma, placing them at increased risk for mental illness. However, this risk can be mitigated by a capacity for resilience, conferred from multiple ecological systems (e.g., family, community), including at an individual biological level. We examined the ability of hair cortisol concentrations and polygenic scores for mental health to predict risk and resilience in a sample of Syrian refugee children (n = 1359). Children were categorized as either at-risk or resilient depending on clinical thresholds for posttraumatic stress disorder, depression, and externalizing behavior problems. Logistic regression was used to examine main and interacting effects while controlling for covariates. Elevated hair cortisol concentrations were significantly associated with reduced resilience (odds ratio (OR)=0.58, 95%CI [0.40, 0.83]) while controlling for levels of war exposure. Polygenic scores for depression, self-harm, and neuroticism were not found to have any significant main effects. However, a significant interaction emerged between hair cortisol and polygenic scores for depression (OR=0.04, 95%CI [0.003 0.47]), suggesting that children predisposed to depression were more at risk for mental health problems when hair cortisol concentrations were high. Our results suggest that biomarkers (separately and in combination) might support early identification of refugee children at risk for mental health problems.

Association of polygenic scores for depression and neuroticism with perceived stress in daily life during a long-lasting stress period

Genetic factors contribute significantly to interindividual differences in the susceptibility to stress-related disorders. As stress can also be conceptualized as environmental exposure, controlled gene-environment interaction (GxE) studies with an in-depth phenotyping may help to unravel mechanisms underlying the interplay between genetic factors and stress. In a prospective-longitudinal quasi-experimental study, we investigated whether polygenic scores (PGS) for depression (DEP-PGS) and neuroticism (NEU-PGS), respectively, were associated with responses to chronic stress in daily life. We examined law students (n = 432) over 13 months. Participants in the stress group experienced a long-lasting stress phase, namely the preparation for the first state examination for law students. The control group consisted of law students without particular stress exposure. In the present manuscript, we analyzed perceived stress levels assessed at high frequency and in an ecologically valid manner by ambulatory assessments as well as depression symptoms and two parameters of the cortisol awakening response. The latter was only assessed in a subsample (n = 196). No associations between the DEP-PGS and stress-related variables were found. However, for the NEU-PGS we found a significant GxE effect. Only in individuals experiencing academic stress a higher PGS for neuroticism predicted stronger increases of perceived stress levels until the exam. At baseline, a higher NEU-PGS was associated with higher perceived stress levels in both groups. Despite the small sample size, we provide preliminary evidence that the genetic disposition for neuroticism is associated with stress level increases in daily life during a long-lasting stress period.

Genetic distance informs polygenic score predictive accuracy

Polygenic scores (PGSs) aggregate the effects of variants across the genome to estimate genetic liability, but have lower performance in external study populations. A new study by Ding et al. has applied a novel framework to estimate the individual-level predictive accuracy of PGSs, and demonstrates that performance reduction occurs linearly with genetic distance.

Potential application of elastic nets for shared polygenicity detection with adapted threshold selection

Current research suggests that hundreds to thousands of single nucleotide polymorphisms (SNPs) with small to modest effect sizes contribute to the genetic basis of many disorders, a phenomenon labeled as polygenicity. Additionally, many such disorders demonstrate polygenic overlap, in which risk alleles are shared at associated genetic loci. A simple strategy to detect polygenic overlap between two phenotypes is based on rank-ordering the univariate p-values from two genome-wide association studies (GWASs). Although high-dimensional variable selection strategies such as Lasso and elastic nets have been utilized in other GWAS analysis settings, they are yet to be utilized for detecting shared polygenicity. In this paper, we illustrate how elastic nets, with polygenic scores as the dependent variable and with appropriate adaptation in selecting the penalty parameter, may be utilized for detecting a subset of SNPs involved in shared polygenicity. We provide theory to better understand our approaches, and illustrate their utility using synthetic datasets. Results from extensive simulations are presented comparing the elastic net approaches with the rank ordering approach, in various scenarios. Results from simulations studies exhibit one of the elastic net approaches to be superior when the correlations among the SNPs are high. Finally, we apply the methods on two real datasets to illustrate further the capabilities, limitations and differences among the methods.

Associations Between Polygenic Scores for Cognitive and Non-cognitive Factors of Educational Attainment and Measures of Behavior, Psychopathology, and Neuroimaging in the Adolescent Brain Cognitive Development Study

Background

Both cognitive and non-cognitive (e.g., traits like curiosity) factors are critical for social and emotional functioning and independently predict educational attainment. These factors are heritable and genetically correlated with a range of health-relevant traits and behaviors in adulthood (e.g., risk-taking, psychopathology). However, whether these associations are present during adolescence, and to what extent these relationships diverge, could have implications for adolescent health and well-being.

Methods

Using data from 5,517 youth of European ancestry from the ongoing Adolescent Brain Cognitive DevelopmentSM Study, we examined associations between polygenic scores (PGS) for cognitive and non-cognitive factors and outcomes related to cognition, socioeconomic status, risk tolerance and decision-making, substance initiation, psychopathology, and brain structure.

Results

Cognitive and non-cognitive PGSs were both positively associated with cognitive performance and family income, and negatively associated with ADHD and severity of psychotic-like experiences. The cognitive PGS was also associated with greater risk-taking, delayed discounting, and anorexia, as well as lower likelihood of nicotine initiation. The cognitive PGS was further associated with cognition scores and anorexia in within-sibling analyses, suggesting these results do not solely reflect the effects of assortative mating or passive gene-environment correlations. The cognitive PGS showed significantly stronger associations with cortical volumes than the non-cognitive PGS and was associated with right hemisphere caudal anterior cingulate and pars-orbitalis in within-sibling analyses, while the non-cognitive PGS showed stronger associations with white matter fractional anisotropy and a significant within-sibling association for right superior corticostriate-frontal cortex.

Conclusions

Our findings suggest that PGSs for cognitive and non-cognitive factors show similar associations with cognition and socioeconomic status as well as other psychosocial outcomes, but distinct associations with regional neural phenotypes in this adolescent sample.

mtPGS: Leverage multiple correlated traits for accurate polygenic score construction

Accurate polygenic scores (PGSs) facilitate the genetic prediction of complex traits and aid in the development of personalized medicine. Here, we develop a statistical method called multi-trait assisted PGS (mtPGS), which can construct accurate PGSs for a target trait of interest by leveraging multiple traits relevant to the target trait. Specifically, mtPGS borrows SNP effect size similarity information between the target trait and its relevant traits to improve the effect size estimation on the target trait, thus achieving accurate PGSs. In the process, mtPGS flexibly models the shared genetic architecture between the target and the relevant traits to achieve robust performance, while explicitly accounting for the environmental covariance among them to accommodate different study designs with various sample overlap patterns. In addition, mtPGS uses only summary statistics as input and relies on a deterministic algorithm with several algebraic techniques for scalable computation. We evaluate the performance of mtPGS through comprehensive simulations and applications to 25 traits in the UK Biobank, where in the real data mtPGS achieves an average of 0.90%-52.91% accuracy gain compared to the state-of-the-art PGS methods. Overall, mtPGS represents an accurate, fast, and robust solution for PGS construction in biobank-scale datasets.

Alcohol use polygenic risk score, social support, and alcohol use among European American and African American adults

Alcohol use is influenced by genetic and environmental factors. We examined the interactive effects between genome-wide polygenic risk scores for alcohol use (alc-PRS) and social support in relation to alcohol use among European American (EA) and African American (AA) adults across sex and developmental stages (emerging adulthood, young adulthood, and middle adulthood). Data were drawn from 4,011 EA and 1,274 AA adults from the Collaborative Study on the Genetics of Alcoholism who were between ages 18-65 and had ever used alcohol. Participants completed the Semi-Structured Assessment for the Genetics of Alcoholism and provided saliva or blood samples for genotyping. Results indicated that social support from friends, but not family, moderated the association between alc-PRS and alcohol use among EAs and AAs (only in middle adulthood for AAs); alc-PRS was associated with higher levels of alcohol use when friend support was low, but not when friend support was high. Associations were similar across sex but differed across developmental stages. Findings support the important role of social support from friends in buffering genetic risk for alcohol use among EA and AA adults and highlight the need to consider developmental changes in the role of social support in relation to alcohol use.

Psychotic experiences and negative symptoms from adolescence to emerging adulthood: developmental trajectories and associations with polygenic scores and childhood characteristics

BACKGROUND

Psychotic experiences and negative symptoms (PENS) are common in non-clinical populations. PENS are associated with adverse outcomes, particularly when they persist. Little is known about the trajectories of PENS dimensions in young people, nor about the precursory factors associated with these trajectories.

METHODS

We conducted growth mixture modelling of paranoia, hallucinations, and negative symptoms across ages 16, 17, and 22 in a community sample (N = 12 049-12 652). We then described the emergent trajectory classes through their associations with genome-wide polygenic scores (GPS) for psychiatric and educational phenotypes, and earlier childhood characteristics.

RESULTS

Three trajectory classes emerged for paranoia, two for hallucinations, and two for negative symptoms. Across PENS, GPS for clinical help-seeking, major depressive disorder, and attention deficit hyperactivity disorder were associated with increased odds of being in the most elevated trajectory class (OR 1.07-1.23). Lower education GPS was associated with the most elevated trajectory class for hallucinations and negative symptoms (OR 0.77-0.91). Conversely for paranoia, higher education GPS was associated with the most elevated trajectory class (OR 1.25). Trajectory class associations were not significant for schizophrenia, obsessive-compulsive disorder, bipolar disorder, or anorexia GPS. Emotional/behaviour problems and life events in childhood were associated with increased odds of being in the most elevated trajectory class across PENS.

CONCLUSIONS

Our results suggest latent heterogeneity in the development of paranoia, hallucinations, and negative symptoms in young people that is associated with specific polygenic scores and childhood characteristics.

Bridging the diversity gap: Analytical and study design considerations for improving the accuracy of trans-ancestry genetic prediction

Genetic prediction of common complex disease risk is an essential component of precision medicine. Currently, genome-wide association studies (GWASs) are mostly composed of European-ancestry samples and resulting polygenic scores (PGSs) have been shown to poorly transfer to other ancestries partly due to heterogeneity of allelic effects between populations. Fixed-effects (FETA) and random-effects (RETA) trans-ancestry meta-analyses do not model such ancestry-related heterogeneity, while ancestry-specific (AS) scores may suffer from low power due to low sample sizes. In contrast, trans-ancestry meta-regression (TAMR) builds ancestry-aware PGS that account for more complex trans-ancestry architectures. Here, we examine the predictive performance of these four PGSs under multiple genetic architectures and ancestry configurations. We show that the predictive performance of FETA and RETA is strongly affected by cross-ancestry genetic heterogeneity, while AS PGS performance decreases in under-represented target populations. TAMR PGS is also impacted by heterogeneity but maintains good prediction performance in most situations, especially in ancestry-diverse scenarios. In simulations of human complex traits, TAMR scores currently explain 25% more phenotypic variance than AS in triglyceride levels and 33% more phenotypic variance than FETA in type 2 diabetes in most non-European populations. Importantly, a high proportion of non-European-ancestry individuals is needed to reach prediction levels that are comparable in those populations to the one observed in European-ancestry studies. Our results highlight the need to rebalance the ancestral composition of GWAS to enable accurate prediction in non-European-ancestry groups, and demonstrate the relevance of meta-regression approaches for compensating some of the current population biases in GWAS.

Filling the gap: Genetic risk assessment in hypercholesterolemia using LDL-C and LPA genetic scores

Routine genetic testing in hypercholesterolemia patients reveals a causative monogenic variant in less than 50% of affected individuals. Incomplete genetic characterization is partly due to polygenic factors influencing low-density-lipoprotein-cholesterol (LDL-C). Additionally, functional variants in the LPA gene affect lipoprotein(a)-associated cholesterol concentrations but are difficult to determine due to the complex structure of the LPA gene. In this study we examined whether complementing standard sequencing with the analysis of genetic scores associated with LDL-C and Lp(a) concentrations improves the diagnostic output in hypercholesterolemia patients. 1.020 individuals including 252 clinically diagnosed hypercholesterolemia patients from the FH Register Austria were analyzed by massive-parallel-sequencing of candidate genes combined with array genotyping, identifying nine novel variants in LDLR. For each individual, validated genetic scores associated with elevated LDL-C and Lp(a) were calculated based on imputed genotypes. Integrating these scores especially the score for Lp(a) increased the proportion of individuals with a clearly defined disease etiology to 68.8% compared to 46.6% in standard genetic testing. The study highlights the major role of Lp(a) in disease etiology in clinically diagnosed hypercholesterolemia patients, of which parts are misclassified. Screening for monogenic causes of hypercholesterolemia and genetic scores for LDL-C and Lp(a) permits more precise diagnosis, allowing individualized treatment.

Genome-wide association studies: utility and limitations for research in physiology

Physiological systems are subject to interindividual variation encoded by genetics. Genome-wide association studies (GWAS) operate by surveying thousands of genetic variants from a substantial number of individuals and assessing their association to a trait of interest, be it a physiological variable, a molecular phenotype (e.g. gene expression), or even a disease or condition. Through a myriad of methods, GWAS downstream analyses then explore the functional consequences of each variant and attempt to ascertain a causal relationship to the phenotype of interest, as well as to delve into its links to other traits. This type of investigation allows mechanistic insights into physiological functions, pathological disturbances and shared biological processes between traits (i.e. pleiotropy). An exciting example is the discovery of a new thyroid hormone transporter (SLC17A4) and hormone metabolising enzyme (AADAT) from a GWAS on free thyroxine levels. Therefore, GWAS have substantially contributed with insights into physiology and have been shown to be useful in unveiling the genetic control underlying complex traits and pathological conditions; they will continue to do so with global collaborations and advances in genotyping technology. Finally, the increasing number of trans-ancestry GWAS and initiatives to include ancestry diversity in genomics will boost the power for discoveries, making them also applicable to non-European populations.

Immunogenetics of lithium response and psychiatric phenotypes in patients with bipolar disorder

The link between bipolar disorder (BP) and immune dysfunction remains controversial. While epidemiological studies have long suggested an association, recent research has found only limited evidence of such a relationship. To clarify this, we investigated the contributions of immune-relevant genetic factors to the response to lithium (Li) treatment and the clinical presentation of BP. First, we assessed the association of a large collection of immune-related genes (4,925) with Li response, defined by the Retrospective Assessment of the Lithium Response Phenotype Scale (Alda scale), and clinical characteristics in patients with BP from the International Consortium on Lithium Genetics (ConLi+Gen, N = 2,374). Second, we calculated here previously published polygenic scores (PGSs) for immune-related traits and evaluated their associations with Li response and clinical features. We found several genes associated with Li response at p < 1×10- 4 values, including HAS3, CNTNAP5 and NFIB. Network and functional enrichment analyses uncovered an overrepresentation of pathways involved in cell adhesion and intercellular communication, which appear to converge on the well-known Li-induced inhibition of GSK-3β. We also found various genes associated with BP's age-at-onset, number of mood episodes, and presence of psychosis, substance abuse and/or suicidal ideation at the exploratory threshold. These included RTN4, XKR4, NRXN1, NRG1/3 and GRK5. Additionally, PGS analyses suggested serum FAS, ECP, TRANCE and cytokine ligands, amongst others, might represent potential circulating biomarkers of Li response and clinical presentation. Taken together, our results support the notion of a relatively weak association between immunity and clinically relevant features of BP at the genetic level.

Structural architecture and brain network efficiency link polygenic scores to intelligence

Intelligence is highly heritable. Genome-wide association studies (GWAS) have shown that thousands of alleles contribute to variation in intelligence with small effect sizes. Polygenic scores (PGS), which combine these effects into one genetic summary measure, are increasingly used to investigate polygenic effects in independent samples. Whereas PGS explain a considerable amount of variance in intelligence, it is largely unknown how brain structure and function mediate this relationship. Here, we show that individuals with higher PGS for educational attainment and intelligence had higher scores on cognitive tests, larger surface area, and more efficient fiber connectivity derived by graph theory. Fiber network efficiency as well as the surface of brain areas partly located in parieto-frontal regions were found to mediate the relationship between PGS and cognitive performance. These findings are a crucial step forward in decoding the neurogenetic underpinnings of intelligence, as they identify specific regional networks that link polygenic predisposition to intelligence.

Joint polygenic and environmental risks for childhood attention-deficit/hyperactivity disorder (ADHD) and ADHD symptom dimensions

Background

attention-deficit/hyperactivity disorder (ADHD) is associated with both polygenic liability and environmental exposures, both intrinsic to the family, such as family conflict, and extrinsic, such as air pollution. However, much less is known about the interplay between environmental and genetic risks relevant to ADHD-a better understanding of which could inform both mechanistic models and clinical prediction algorithms.

Methods

Two independent data sets, the population-based Adolescent Brain Cognitive Development Study (ABCD) (N = 11,876) and the case-control Oregon-ADHD-1000 (N = 1449), were used to examine additive (G + E) and interactive (GxE) effects of selected polygenic risk scores (PRS) and environmental factors in a cross-sectional design. Genetic risk was measured using PRS for nine mental health disorders/traits. Exposures included family income, family conflict/negative sentiment, and geocoded measures of area deprivation, lead exposure risk, and air pollution exposure (nitrogen dioxide and fine particulate matter).

Results

ADHD PRS and family conflict jointly predicted concurrent ADHD symptoms in both cohorts. Additive-effects models, including both genetic and environmental factors, explained significantly more variation in symptoms than any individual factor alone (joint R 2 = .091 for total symptoms in ABCD; joint R 2 = .173 in Oregon-ADHD-1000; all delta-R 2 p-values <2e-7). Significant effect size heterogeneity across ancestry groups was observed for genetic and environmental factors (e.g., Q = 9.01, p = .011 for major depressive disorder PRS; Q = 13.34, p = .001 for area deprivation). GxE interactions observed in the full ABCD cohort suggested stronger environmental effects when genetic risk is low, though they did not replicate.

Conclusions

Reproducible additive effects of PRS and family environment on ADHD symptoms were found, but GxE interaction effects were not replicated and appeared confounded by ancestry. Results highlight the potential value of combining exposures and PRS in clinical prediction algorithms. The observed differences in risks across ancestry groups warrant further study to avoid health care disparities.

Sub-types of insomnia in adolescents: Insights from a quantitative/molecular twin study

Background

Insomnia with short sleep duration has been postulated as more severe than that accompanied by normal/long sleep length. While the short duration subtype is considered to have greater genetic influence than the other subtype, no studies have addressed this question. This study aimed to compare these subtypes in terms of: (1) the heritability of insomnia symptoms; (2) polygenic scores (PGS) for insomnia symptoms and sleep duration; (3) the associations between insomnia symptoms and a wide variety of traits/disorders.

Methods

The sample comprised 4000 pairs of twins aged 16 from the Twins Early Development Study. Twin models were fitted to estimate the heritability of insomnia in both groups. PGS were calculated for self-reported insomnia and sleep duration and compared among participants with short and normal/long sleep duration.

Results

Heritability was not significantly different in the short sleep duration group (A = 0.13 [95%CI = 0.01, 0.32]) and the normal/long sleep duration group (A = 0.35 [95%CI = 0.29, 0.40]). Shared environmental factors accounted for a substantial proportion of the variance in the short sleep duration group (C = 0.19 [95%CI = 0.05, 0.32]) but not in the normal/long sleep duration group (C = 0.00 [95%CI = 0.00, 0.04]). PGS did not differ significantly between groups although results were in the direction expected by the theory. Our results also showed that insomnia with short (as compared to normal/long) sleep duration had a stronger association with anxiety and depression (p < .05)-although not once adjusting for multiple testing.

Conclusions

We found mixed results in relation to the expected differences between the insomnia subtypes in adolescents. Future research needs to further establish cut-offs for 'short' sleep at different developmental stages and employ objective measures of sleep.

Predicting eating disorder and anxiety symptoms using disorder-specific and transdiagnostic polygenic scores for anorexia nervosa and obsessive-compulsive disorder

BACKGROUND

Clinical, epidemiological, and genetic findings support an overlap between eating disorders, obsessive-compulsive disorder (OCD), and anxiety symptoms. However, little research has examined the role of genetics in the expression of underlying phenotypes. We investigated whether the anorexia nervosa (AN), OCD, or AN/OCD transdiagnostic polygenic scores (PGS) predict eating disorder, OCD, and anxiety symptoms in a large developmental cohort in a sex-specific manner.

METHODS

Using summary statistics from Psychiatric Genomics Consortium AN and OCD genome-wide association studies, we conducted an AN/OCD transdiagnostic genome-wide association meta-analysis. We then calculated AN, OCD, and AN/OCD PGS in participants from the Avon Longitudinal Study of Parents and Children to predict eating disorder, OCD, and anxiety symptoms, stratified by sex (combined N = 3212-5369 per phenotype).

RESULTS

The PGS prediction of eating disorder, OCD, and anxiety phenotypes differed between sexes, although effect sizes were small. AN and AN/OCD PGS played a more prominent role in predicting eating disorder and anxiety risk than OCD PGS, especially in girls. AN/OCD PGS provided a small boost over AN PGS in the prediction of some anxiety symptoms. All three PGS predicted higher compulsive exercise across different developmental timepoints [β = 0.03 (s.e. = 0.01) for AN and AN/OCD PGS at age 14; β = 0.05 (s.e. = 0.02) for OCD PGS at age 16] in girls.

CONCLUSIONS

Compulsive exercise may have a transdiagnostic genetic etiology, and AN genetic risk may play a role in the presence of anxiety symptoms. Converging with prior twin literature, our results also suggest that some of the contribution of genetic risk may be sex-specific.

Low and differential polygenic score generalizability among African populations due largely to genetic diversity

African populations are vastly underrepresented in genetic studies but have the most genetic variation and face wide-ranging environmental exposures globally. Because systematic evaluations of genetic prediction had not yet been conducted in ancestries that span African diversity, we calculated polygenic risk scores (PRSs) in simulations across Africa and in empirical data from South Africa, Uganda, and the United Kingdom to better understand the generalizability of genetic studies. PRS accuracy improves with ancestry-matched discovery cohorts more than from ancestry-mismatched studies. Within ancestrally and ethnically diverse South African individuals, we find that PRS accuracy is low for all traits but varies across groups. Differences in African ancestries contribute more to variability in PRS accuracy than other large cohort differences considered between individuals in the United Kingdom versus Uganda. We computed PRS in African ancestry populations using existing European-only versus ancestrally diverse genetic studies; the increased diversity produced the largest accuracy gains for hemoglobin concentration and white blood cell count, reflecting large-effect ancestry-enriched variants in genes known to influence sickle cell anemia and the allergic response, respectively. Differences in PRS accuracy across African ancestries originating from diverse regions are as large as across out-of-Africa continental ancestries, requiring commensurate nuance.

Do Children Think it is Important to Predict Learning and Behaviour Problems, and Do They Think Genetic Screening Has a Role to Play in This?

This study explores how capable young children are of thinking about a potential future that uses DNA screening to assess an individual's likelihood of experiencing learning or behaviour difficulties. Puppets and a scenario-based approach were used to ask children aged 4-10 (n = 165) whether they thought DNA screening might be helpful or harmful. A content analysis derived six categories: (1) 'Worried about being - and being seen as - different'; (2) 'Beliefs about the origins of learning and behaviour'; (3) 'Testing is harmful'; (4) 'Testing could help'; (5) 'How soon is too soon for testing?'; and (6) 'What's the point?'. Findings indicate young children, as key stakeholders, can make useful contributions to public debate in this important and controversial area.

The production of within-family inequality: Insights and implications of integrating genetic data

The integration of genetic data within large-scale social and health surveys provides new opportunities to test long-standing theories of parental investments in children and within-family inequality. Genetic predictors, called polygenic scores, allow novel assessments of young children's abilities that are uncontaminated by parental investments, and family-based samples allow indirect tests of whether children's abilities are reinforced or compensated. We use over 16,000 sibling pairs from the UK Biobank to test whether the relative ranking of siblings' polygenic scores for educational attainment is consequential for actual attainments. We find evidence consistent with compensatory processes, on average, where the association between genotype and phenotype of educational attainment is reduced by over 20% for the higher-ranked sibling compared to the lower-ranked sibling. These effects are most pronounced in high socioeconomic status areas. We find no evidence that similar processes hold in the case of height or for relatives who are not full biological siblings (e.g. cousins). Our results provide a new use of polygenic scores to understand processes that generate within-family inequalities and also suggest important caveats to causal interpretations the effects of polygenic scores using sibling difference designs. Future work should seek to replicate these findings in other data and contexts.

Association of parental divorce, discord, and polygenic risk with children's alcohol initiation and lifetime risk for alcohol use disorder

BACKGROUND

Parental divorce and discord are associated with poorer alcohol-related outcomes for offspring. However, not all children exposed to these stressors develop alcohol problems. Our objective was to test gene-by-environment interaction effects whereby children's genetic risk for alcohol problems modifies the effects of parental divorce and discord to predict alcohol outcomes.

METHODS

The sample included European (EA; N = 5608, 47% male, Mage  ~ 36 years) and African (AA; N = 1714, 46% female, Mage  ~ 33 years) ancestry participants from the Collaborative Study on the Genetics of Alcoholism. Outcomes included age at initiation of regular drinking and lifetime DSM-5 alcohol use disorder (AUD). Predictors included parental divorce, parental relationship discord, and offspring alcohol problems polygenic risk scores (PRSALC ). Mixed effects Cox proportional hazard models were used to examine alcohol initiation and generalized linear mixed effects models were used to examine lifetime AUD. Tests of PRS moderation of the effects of parental divorce/relationship discord on alcohol outcomes were examined on multiplicative and additive scales.

RESULTS

Among EA participants, parental divorce, parental discord, and higher PRSALC were associated with earlier alcohol initiation and greater lifetime AUD risk. Among AA participants, parental divorce was associated with earlier alcohol initiation and discord was associated with earlier initiation and AUD. PRSALC was not associated with either. Parental divorce/discord and PRSALC interacted on an additive scale in the EA sample, but no interactions were found in AA participants.

CONCLUSIONS

Children's genetic risk for alcohol problems modifies the impact of parental divorce/discord, consistent with an additive model of diathesis-stress interaction, with some differences across ancestry.

The interactive association of adverse childhood experiences and polygenic susceptibility with depressive symptoms and chronic inflammation in older adults: a prospective cohort study

BACKGROUND

Adverse childhood experiences (ACEs) and genetic liability are important risk factors for depression and inflammation. However, little is known about the gene-environment (G × E) mechanisms underlying their aetiology. For the first time, we tested the independent and interactive associations of ACEs and polygenic scores of major depressive disorder (MDD-PGS) and C-reactive protein (CRP-PGS) with longitudinal trajectories of depression and chronic inflammation in older adults.

METHODS

Data were drawn from the English longitudinal study of ageing (N~3400). Retrospective information on ACEs was collected in wave3 (2006/07). We calculated a cumulative risk score of ACEs and also assessed distinct dimensions separately. Depressive symptoms were ascertained on eight occasions, from wave1 (2002/03) to wave8 (2016/17). CRP was measured in wave2 (2004/05), wave4 (2008/09), and wave6 (2012/13). The associations of the risk factors with group-based depressive-symptom trajectories and repeated exposure to high CRP (i.e. ⩾3 mg/L) were tested using multinomial and ordinal logistic regression.

RESULTS

All types of ACEs were independently associated with high depressive-symptom trajectories (OR 1.44, 95% CI 1.30-1.60) and inflammation (OR 1.08, 95% CI 1.07-1.09). The risk of high depressive-symptom trajectories (OR 1.47, 95% CI 1.28-1.70) and inflammation (OR 1.03, 95% CI 1.01-1.04) was also higher for participants with higher MDD-PGS. G×E analyses revealed that the associations between ACEs and depressive symptoms were larger among participants with higher MDD-PGS (OR 1.13, 95% CI 1.04-1.23). ACEs were also more strongly related to inflammation in participants with higher CRP-PGS (OR 1.02, 95% CI 1.01-1.03).

CONCLUSIONS

ACEs and polygenic susceptibility were independently and interactively associated with elevated depressive symptoms and chronic inflammation, highlighting the clinical importance of assessing both ACEs and genetic risk factors to design more targeted interventions.

Disentangling the influences of parental genetics on offspring's cognition, education, and psychopathology via genetic and phenotypic pathways

BACKGROUND

Specific pathways of intergenerational transmission of behavioral traits remain unclear. Here, we aim to investigate how parental genetics influence offspring cognition, educational attainment, and psychopathology in youth.

METHODS

Participants for the discovery sample were 2,189 offspring (aged 6-14 years), 1898 mothers and 1,017 fathers who underwent genotyping, psychiatric, and cognitive assessments. We calculated polygenic scores (PGS) for cognition, educational attainment, attention-deficit hyperactivity disorder (ADHD), and schizophrenia for the trios. Phenotypes studied included educational and cognitive measures, ADHD and psychotic symptoms. We used a stepwise approach and multiple mediation models to analyze the effect of parental PGS on offspring traits via offspring PGS and parental phenotype. Significant results were replicated in a sample of 1,029 adolescents, 363 mothers, and 307 fathers.

RESULTS

Maternal and paternal PGS for cognition influenced offspring general intelligence and executive function via offspring PGS (genetic pathway) and parental education (phenotypic pathway). Similar results were found for parental PGS for educational attainment and offspring reading and writing skills. These pathways fully explained associations between parental PGS and offspring phenotypes, without residual direct association. Associations with maternal, but not paternal, PGS were replicated. No associations were found between parental PGS for psychopathology and offspring specific symptoms.

CONCLUSIONS

Our findings indicate that parental genetics influences offspring cognition and educational attainment by genetic and phenotypic pathways, suggesting the expression of parental phenotypes partially explain the association between parental genetic risk and offspring outcomes. Multiple mediations might represent an effective approach to disentangle distinct pathways for intergenerational transmission of behavioral traits.

Beyond orchids and dandelions: Susceptibility to environmental influences is not bimodal

This study focused on generality versus specificity of susceptibility of effects of eight family and child-care exposures measured between 3 and 54 months of age (e.g., sensitive parenting, child-care quality) on five child development outcomes assessed at age 4.5 years (e.g. behavior problems, preacademic skill), using data from The National Institute of Child Health and Human Development (NICHD) Study of Early Child Care and Youth Development (n = 1,364, boys = 705; White = 1,097, Black = 176, other = 91), while applying a novel influence-statistics method. Results indicated that susceptibility across the environment-predictor:child-outcome associations is normally rather than bimodally (i.e., orchid-dandelion) distributed. Analysis of susceptibility documents both domain generality and specificity of developmental plasticity, with effect sizes proving small in the former case. As predicted, children who as infants had difficult temperaments or who scored higher on a polygenic-plasticity score (serotonin-transporter-linked promoter region [5-HTTLPR], dopamine receptor D4 [DRD4], brain-derived neurotrophic factor [BDNF]) proved somewhat more susceptible to some of the environmental effects investigated. Results lead to the recommendation that two-types-of-individuals vis-a-vis susceptibility to environmental influences be questioned and general-trait conceptions of susceptibility be further investigated.

Human height: a model common complex trait

CONTEXT

Like other complex phenotypes, human height reflects a combination of environmental and genetic factors, but is notable for being exceptionally easy to measure. Height has therefore been commonly used to make observations later generalised to other phenotypes though the appropriateness of such generalisations is not always considered.

OBJECTIVES

We aimed to assess height's suitability as a model for other complex phenotypes and review recent advances in height genetics with regard to their implications for complex phenotypes more broadly.

METHODS

We conducted a comprehensive literature search in PubMed and Google Scholar for articles relevant to the genetics of height and its comparatibility to other phenotypes.

RESULTS

Height is broadly similar to other phenotypes apart from its high heritability and ease of measurment. Recent genome-wide association studies (GWAS) have identified over 12,000 independent signals associated with height and saturated height's common single nucleotide polymorphism based heritability of height within a subset of the genome in individuals similar to European reference populations.

CONCLUSIONS

Given the similarity of height to other complex traits, the saturation of GWAS's ability to discover additional height-associated variants signals potential limitations to the omnigenic model of complex-phenotype inheritance, indicating the likely future power of polygenic scores and risk scores, and highlights the increasing need for large-scale variant-to-gene mapping efforts.

Integrating Common Risk Factors with Polygenic Scores Improves the Prediction of Type 2 Diabetes

We tested associations between 13 established genetic variants and type 2 diabetes (T2D) in 1371 study participants from the Volga-Ural region of the Eurasian continent, and evaluated the predictive ability of the model containing polygenic scores for the variants associated with T2D in our dataset, alone and in combination with other risk factors such as age and sex. Using logistic regression analysis, we found associations with T2D for the CCL20 rs6749704 (OR = 1.68, P_FDR = 3.40 × 10^-5), CCR5 rs333 (OR = 1.99, P_FDR = 0.033), ADIPOQ rs17366743 (OR = 3.17, P_FDR = 2.64 × 10^-4), TCF7L2 rs114758349 (OR = 1.77, P_FDR = 9.37 × 10^-5), and CCL2 rs1024611 (OR = 1.38, P_FDR = 0.033) polymorphisms. We showed that the most informative prognostic model included weighted polygenic scores for these five loci, and non-genetic factors such as age and sex (AUC 85.8%, 95%CI 83.7-87.8%). Compared to the model containing only non-genetic parameters, adding the polygenic score for the five T2D-associated loci showed improved net reclassification (NRI = 37.62%, 1.39 × 10^-6). Inclusion of all 13 tested SNPs to the model with age and sex did not improve the predictive ability compared to the model containing five T2D-associated variants (NRI = -17.86, p = 0.093). The five variants associated with T2D in people from the Volga-Ural region are linked to inflammation (CCR5, CCL2, CCL20) and glucose metabolism regulation (TCF7L, ADIPOQ2). Further studies in independent groups of T2D patients should validate the prognostic value of the model and elucidate the molecular mechanisms of the disease development.

Multivariate analyses of molecular genetic associations between childhood psychopathology and adult mood disorders and related traits

Ubiquitous associations have been detected between different types of childhood psychopathology and polygenic risk scores based on adult psychiatric disorders and related adult outcomes, indicating that genetic factors partly explain the association between childhood psychopathology and adult outcomes. However, these analyses in general do not take into account the correlations between the adult trait and disorder polygenic risk scores. This study aimed to further clarify the influence of genetic factors on associations between childhood psychopathology and adult outcomes by accounting for these correlations. Using a multivariate multivariable regression, we analyzed associations of childhood attention-deficit/hyperactivity disorder (ADHD), internalizing, and social problems, with polygenic scores (PGS) of adult disorders and traits including major depression, bipolar disorder, subjective well-being, neuroticism, insomnia, educational attainment, and body mass index (BMI), derived for 20,539 children aged 8.5-10.5 years. After correcting for correlations between the adult phenotypes, major depression PGS were associated with all three childhood traits, that is, ADHD, internalizing, and social problems. In addition, BMI PGS were associated with ADHD symptoms and social problems, while neuroticism PGS were only associated with internalizing problems and educational attainment PGS were only associated with ADHD symptoms. PGS of bipolar disorder, subjective well-being, and insomnia were not associated with any childhood traits. Our findings suggest that associations between childhood psychopathology and adult traits like insomnia and subjective well-being may be primarily driven by genetic factors that influence adult major depression. Additionally, specific childhood phenotypes are genetically associated with educational attainment, BMI and neuroticism.

Predictive models for abdominal aortic aneurysms using polygenic scores and PheWAS-derived risk factors

Abdominal aortic aneurysms (AAA) are common enlargements of the abdominal aorta which can grow larger until rupture, often leading to death. Detection of AAA is often by ultrasonography and screening recommendations are mostly directed at men over 65 with a smoking history. Recent large-scale genome-wide association studies have identified genetic loci associated with AAA risk. We combined known risk factors, polygenic risk scores (PRS) and precedent clinical diagnoses from electronic health records (EHR) to develop predictive models for AAA, and compared performance against screening recommendations. The PRS included genome-wide summary statistics from the Million Veteran Program and FinnGen (10,467 cases, 378,713 controls of European ancestry), with optimization in Vanderbilt's BioVU and validated in the eMERGE Network, separately across both White and Black participants. Candidate diagnoses were identified through a temporally-oriented Phenome-wide association study in independent EHR data from Vanderbilt, and features were selected via elastic net. We calculated C-statistics in eMERGE for models including PRS, phecodes, and covariates using regression weights from BioVU. The AUC for the full model in the test set was 0.883 (95% CI 0.873-0.892), 0.844 (0.836-0.851) for covariates only, 0.613 (95% CI 0.604-0.622) when using primary USPSTF screening criteria, and 0.632 (95% CI 0.623-0.642) using primary and secondary criteria. Brier scores were between 0.003 and 0.023 for our models indicating good calibration, and net reclassification improvement over combined primary and secondary USPSTF criteria was 0.36-0.60. We provide PRS for AAA which are strongly associated with AAA risk and add to predictive model performance. These models substantially improve identification of people at risk of a AAA diagnosis compared with existing guidelines, with evidence of potential applicability in minority populations.

The next 10 years of behavioural genomic research

Background

The explosion caused by the fusion of quantitative genetics and molecular genetics will transform behavioural genetic research in child and adolescent psychology and psychiatry.

Methods

Although the fallout has not yet settled, the goal of this paper is to predict the next 10 years of research in what could be called behavioural genomics.

Results

I focus on three research directions: the genetic architecture of psychopathology, causal modelling of gene-environment interplay, and the use of DNA as an early warning system.

Conclusion

Eventually, whole-genome sequencing will be available for all newborns, which means that behavioural genomics could potentially be applied ubiquitously in research and clinical practice.

Race, polygenic risk and their association with incident dementia among older US adults

Dementia incidence increases steadily with age at rates that may vary across racial groups. This racial disparity may be attributable to polygenic risk, as well as lifestyle and behavioural factors. We examined whether Alzheimer's disease polygenic score and race predict Alzheimer's disease and other related dementia incidence differentially by sex and mediation through polygenic scores for other health and behavioural conditions. We used longitudinal data from the nationally representative Health and Retirement Study. We restricted participants to those with complete data on 31 polygenic scores, including Alzheimer's disease polygenic score (2006-2012). Among participants aged 55 years and older in 2008, we excluded those with any memory problems between 2006 and 2008 and included those with complete follow-up on incident Alzheimer's disease and all-cause dementia, between 2010 and 2018 (N = 9683), based on self- or proxy-diagnosis every 2 years (2010, 2012, 2014, 2016 and 2018). Cox proportional hazards and 4-way decomposition models were conducted. Analyses were also stratified by sex and by race. There were racial differences in all-cause dementia incidence (age and sex-adjusted model, per standard deviation: hazard ratio, HR = 1.34, 95% confidence interval, CI: 1.09-1.65, P = 0.007), partially driven by educational attainment and income. We also found independent associations of race (age and sex-adjusted model, African American versus White adults: HR = 2.07, 95% CI: 1.52-2.83, P < 0.001) and Alzheimer's disease polygenic score (age and sex-adjusted model, per SD: HR = 1.37, 95% CI: 1.00-1.87, P < 0.001) with Alzheimer's disease incidence, including sex differences whereby women had a stronger effect of Alzheimer's disease polygenic score on Alzheimer's disease incidence compared with men (P < 0.05 for sex by Alzheimer's disease polygenic score interaction) adjusting for race and other covariates. The total impact of Alzheimer's disease polygenic scores on Alzheimer's disease incidence was mostly direct, while the effect of race on all-cause dementia incidence was mediated through socio-economic, lifestyle and health-related factors. Finally, among the 30 polygenic scores we examined, the total effects on the pathway Alzheimer's disease polygenic score --> Other polygenic score --> Incident Alzheimer's or all-cause dementia, were statistically significant for all, driven primarily by the controlled direct effect (P< 0. 001). In conclusion, both race and Alzheimer's disease polygenic scores were associated independently with Alzheimer's disease and all-cause dementia incidence. Alzheimer's disease polygenic score was more strongly linked to incident Alzheimer's disease among women, while racial difference in all-cause dementia was explained by other factors including socio-economic status.

Gattaca as a lens on contemporary genetics: Marking 25 years into the film's "not-too-distant" future

The 1997 film Gattaca has emerged as a canonical pop culture reference used to discuss modern controversies in genetics and bioethics. It appeared in theaters a few years prior to the announcement of the "completion" of the human genome (2000), as the science of human genetics was developing a renewed sense of its social implications. The story is set in a near-future world in which parents can, with technological assistance, influence the genetic composition of their offspring on the basis of predicted life outcomes. The current moment-25 years after the film's release-offers an opportunity to reflect on where society currently stands with respect to the ideas explored in Gattaca. Here, we review and discuss several active areas of genetic research-genetic prediction, embryo selection, forensic genetics, and others-that interface directly with scenes and concepts in the film. On its silver anniversary, we argue that Gattaca remains an important reflection of society's expectations and fears with respect to the ways that genetic science has manifested in the real world. In accompanying supplemental material, we offer some thought questions to guide group discussions inside and outside of the classroom.

ExPRSweb: An online repository with polygenic risk scores for common health-related exposures

Complex traits are influenced by genetic risk factors, lifestyle, and environmental variables, so-called exposures. Some exposures, e.g., smoking or lipid levels, have common genetic modifiers identified in genome-wide association studies. Because measurements are often unfeasible, exposure polygenic risk scores (ExPRSs) offer an alternative to study the influence of exposures on various phenotypes. Here, we collected publicly available summary statistics for 28 exposures and applied four common PRS methods to generate ExPRSs in two large biobanks: the Michigan Genomics Initiative and the UK Biobank. We established ExPRSs for 27 exposures and demonstrated their applicability in phenome-wide association studies and as predictors for common chronic conditions. Especially the addition of multiple ExPRSs showed, for several chronic conditions, an improvement compared to prediction models that only included traditional, disease-focused PRSs. To facilitate follow-up studies, we share all ExPRS constructs and generated results via an online repository called ExPRSweb.

Educational Attainment Polygenic Scores: Examining Evidence for Gene-Environment Interplay with Adolescent Alcohol, Tobacco and Cannabis Use

Genes associated with educational attainment may be related to or interact with adolescent alcohol, tobacco and cannabis use. Potential gene-environment interplay between educational attainment polygenic scores (EA-PGS) and adolescent alcohol, tobacco, and cannabis use was evaluated with a series of regression models fitted to data from a sample of 1871 adult Australian twins. All models controlled for age, age², cohort, sex and genetic ancestry as fixed effects, and a genetic relatedness matrix was included as a random effect. Although there was no evidence that adolescent alcohol, tobacco or cannabis use interacted with EA-PGS to influence educational attainment, there was a significant, positive gene-environment correlation with adolescent alcohol use at all PGS thresholds (ps <.02). Higher EA-PGS were associated with an increased likelihood of using alcohol as an adolescent (ΔR² ranged from 0.5% to 1.1%). The positive gene-environment correlation suggests a complex relationship between educational attainment and alcohol use that is due to common genetic factors.

Commentary: Will genomics revolutionise research on gene-environment interplay?

The synthesis of quantitative genetics and molecular genetics is transforming research in the behavioural sciences. The ability to measure inherited DNA differences directly has led to polygenic scores and to new methods to estimate heritability and genetic correlations. This issue provides examples of how these advances can be appllied to research on gene-environment interplay in developmental psychopathology.

A Shifting Relationship Between Sex Hormone-Binding Globulin and Total Testosterone Across Puberty in Boys

CONTEXT

Sex hormone-binding globulin (SHBG) is associated with levels of total testosterone (total-T), and both total-T and SHBG are associated with obesity.

OBJECTIVE

We aimed to clarify the nature of the relationship between testosterone and SHBG and improve our understanding of their relationships with obesity. We hypothesize that the hypothalamic-pituitary-gonadal axis contributes to the homeostasis of testosterone by increasing the production of gonadal testosterone through a feedback mechanism that might operate differently at different pubertal stages.

METHODS

We investigated the dynamics of the relationship between SHBG, total-T, and body mass index (BMI) throughout puberty (from age 9 to 17) using longitudinal data obtained in 507 males. The directionality of this relationship was explored using polygenic scores of SHBG and total-T, and a two-sample Mendelian Randomization (MR) in male adults.

RESULTS

Consistent with our hypothesis, we found positive relationships between SHBG and total-T at age 15 and 17 but either no relationship or a negative relationship during the earlier time points. Such shifting relationships explained age-related changes in the association between total-T and BMI. Polygenic scores of SHBG and total-T in mediation analyses and the two-sample MR in male adults suggested an effect of SHBG on total-T but also a somewhat weaker effect of total-T on SHBG. Two-sample MR also showed an effect of BMI on SHBG but no effect of SHBG on BMI.

CONCLUSION

These results clarify the nature of the relationship between testosterone and SHBG during puberty and adulthood and shed new light on their possible relationship with obesity.

Gene-environment interaction using polygenic scores: Do polygenic scores for psychopathology moderate predictions from environmental risk to behavior problems?

The DNA revolution has energized research on interactions between genes and environments (GxE) by creating indices of G (polygenic scores) that are powerful predictors of behavioral traits. Here, we test the extent to which polygenic scores for attention-deficit/hyperactivity disorder and neuroticism moderate associations between parent reports of their children's environmental risk (E) at ages 3 and 4 and teacher ratings of behavior problems (hyperactivity/inattention, conduct problems, emotional symptoms, and peer relationship problems) at ages 7, 9 and 12. The sampling frame included up to 6687 twins from the Twins Early Development Study. Our analyses focused on relative effect sizes of G, E and GxE in predicting behavior problems. G, E and GxE predicted up to 2%, 2% and 0.4%, respectively, of the variance in externalizing behavior problems (hyperactivity/inattention and conduct problems) across ages 7, 9 and 12, with no clear developmental trends. G and E predictions of emotional symptoms and peer relationship problems were weaker. A quarter (12 of 48) of our tests of GxE were nominally significant (p = .05). Increasing the predictive power of G and E would enhance the search for GxE.

Genetic assortative mating for schizophrenia and bipolar disorder

BACKGROUND

Psychiatric disorders are highly polygenic and show patterns of partner resemblance. Partner resemblance has direct population-level genetic implications if it is caused by assortative mating, but not if it is caused by convergence or social homogamy. Using genetics may help distinguish these different mechanisms. Here, we investigated whether partner resemblance for schizophrenia and bipolar disorder is influenced by assortative mating using polygenic risk scores (PRSs).

METHODS

PRSs from The Danish High-Risk and Resilience Study-VIA 7 were compared between parents in three subsamples: population-based control parent pairs (N=198), parent pairs where at least one parent had schizophrenia (N=193), and parent pairs where at least one parent had bipolar disorder (N=115).

RESULTS

The PRS for schizophrenia was predictive of schizophrenia in the full sample and showed a significant correlation between parent pairs (r=0.121, p=0.0440), indicative of assortative mating. The PRS for bipolar disorder was also correlated between parent pairs (r=0.162, p=0.0067), but it was not predictive of bipolar disorder in the full sample, limiting the interpretation.

CONCLUSIONS

Our study provides genetic evidence for assortative mating for schizophrenia, with important implications for our understanding of the genetics of schizophrenia.

The pharmacogenomics of selective serotonin reuptake inhibitors

Antidepressant medications are frequently used as the first line of treatment for depression. However, their effectiveness is highly variable and influenced by genetic factors. Recently, pharmacogenetic studies, including candidate-gene, genome-wide association studies or polygenic risk scores, have attempted to uncover the genetic architecture of antidepressant response. Genetic variants in at least 27 genes are linked to antidepressant treatment response in both coding and non-coding genomic regions, but evidence is largely inconclusive due to the high polygenicity of the trait and limited cohort sizes in published studies. Future studies should increase the number and diversity of participants to yield sufficient statistical power to characterize the genetic underpinnings and biological mechanisms of treatment response, improve results generalizability and reduce racial health-related inequities.

Using DNA to predict behaviour problems from preschool to adulthood

BACKGROUND

One goal of the DNA revolution is to predict problems in order to prevent them. We tested here if the prediction of behaviour problems from genome-wide polygenic scores (GPS) can be improved by creating composites across ages and across raters and by using a multi-GPS approach that includes GPS for adult psychiatric disorders as well as for childhood behaviour problems.

METHOD

Our sample included 3,065 genotyped unrelated individuals from the Twins Early Development Study who were assessed longitudinally for hyperactivity, conduct, emotional problems, and peer problems as rated by parents, teachers, and children themselves. GPS created from 15 genome-wide association studies were used separately and jointly to test the prediction of behaviour problems composites (general behaviour problems, externalising, and internalising) across ages (from age 2 to 21) and across raters in penalised regression models. Based on the regression weights, we created multi-trait GPS reflecting the best prediction of behaviour problems. We compared GPS prediction to twin heritability using the same sample and measures.

RESULTS

Multi-GPS prediction of behaviour problems increased from <2% of the variance for observed traits to up to 6% for cross-age and cross-rater composites. Twin study estimates of heritability, although to a lesser extent, mirrored patterns of multi-GPS prediction as they increased from <40% to 83%.

CONCLUSIONS

The ability of GPS to predict behaviour problems can be improved by using multiple GPS, cross-age composites and cross-rater composites, although the effect sizes remain modest, up to 6%. Our approach can be used in any genotyped sample to create multi-trait GPS predictors of behaviour problems that will be more predictive than polygenic scores based on a single age, rater, or GPS.