Population-specific reference panels are crucial for genetic analyses: an example of the CREBRF locus in Native Hawaiians

The accuracy of polygenic score models for BMI and Type II diabetes in the Native Hawaiian population

Polygenic scores (PGS) are promising in stratifying individuals based on the genetic susceptibility to complex diseases or traits. However, the accuracy of PGS models, typically trained in European- or East Asian-ancestry populations, tend to perform poorly in other ethnic minority populations and their accuracies have not been evaluated for Native Hawaiians. In particular, for body mass index (BMI) and type-2 diabetes (T2D), Polynesian-ancestry individuals such as Native Hawaiians or Samoans exhibit varied distribution from other continental populations, but are understudied, particularly in the context of PGS. Using BMI and T2D as examples of metabolic traits of importance to Polynesian populations (along with height as a comparison of a similarly highly polygenic trait), here we examine the prediction accuracies of PGS models in a large Native Hawaiian sample from the Multiethnic Cohort with up to 5300 individuals. We find evidence of lowered prediction accuracies for the PGS models in some cases, particularly for height. We also find that using the Native Hawaiian samples as an optimization cohort during training does not consistently improve PGS performance. Moreover, even the best-performing PGS models among Native Hawaiians have lowered prediction accuracy among the subset of individuals most enriched with Polynesian ancestry. Our findings indicate that factors such as admixture histories, sample size, and diversity in GWAS can influence PGS performance for complex traits among Native Hawaiian samples. This study provides an initial survey of PGS performance among Native Hawaiians and exposes the current gaps and challenges associated with improving polygenic prediction models for underrepresented minority populations.

A likelihood-based framework for demographic inference from genealogical trees

The demographic history of a population underlies patterns of genetic variation and is encoded in the gene-genealogical trees of the sampled haplotypes. Here we propose a demographic inference framework called the genealogical likelihood (gLike). Our method uses a graph-based structure to summarize the relationships among all lineages in a gene-genealogical tree with all possible trajectories of population memberships through time and derives the full likelihood across trees under a parameterized demographic model. We show through simulations and empirical applications that for populations that have experienced multiple admixtures, gLike can accurately estimate dozens of demographic parameters, including ancestral population sizes, admixture timing and admixture proportions, and it outperforms conventional demographic inference methods using the site frequency spectrum. Taken together, our proposed gLike framework harnesses underused genealogical information to offer high sensitivity and accuracy in inferring complex demographies for humans and other species.

Native Hawaiian and Pacific Islander populations in genomic research

The role of genomic research and medicine in improving health continues to grow significantly, highlighting the need for increased equitable inclusion of diverse populations in genomics. Native Hawaiian and Pacific Islander (NHPI) communities are often missing from these efforts to ensure that the benefits of genomics are accessible to all individuals. In this article, we analyze the qualities of NHPI populations relevant to their inclusion in genomic research and investigate their current representation using data from the genome-wide association studies (GWAS) catalog. A discussion of the barriers NHPI experience regarding participating in research and recommendations to improve NHPI representation in genomic research are also included.

Associations between fasting glucose rate-of-change and the missense variant, rs373863828, in an adult Samoan cohort

BACKGROUND

The A allele of rs373863828 in CREB3 regulatory factor is associated with high Body Mass Index, but lower odds of type 2 diabetes. These associations have been replicated elsewhere, but to date all studies have been cross-sectional. Our aims were (1) to describe the development of type 2 diabetes and change in fasting glucose between 2010 and 2018 among a longitudinal cohort of adult Samoans without type 2 diabetes or who were not using diabetes medications at baseline, and (2) to examine associations between fasting glucose rate-of-change (mmol/L per year) and the A allele of rs373863828.

METHODS

We describe and test differences in fasting glucose, the development of type 2 diabetes, body mass index, age, smoking status, physical activity, urbanicity of residence, and household asset scores between 2010 and 2018 among a cohort of n = 401 adult Samoans, selected to have a ~2:2:1 ratio of GG:AG: AA rs373863828 genotypes. Multivariate linear regression was used to test whether fasting glucose rate-of-change was associated with rs373863828 genotype, and other baseline variables.

RESULTS

By 2018, fasting glucose and BMI significantly increased among all genotype groups, and a substantial portion of the sample developed type 2 diabetes mellitus. The A allele was associated with a lower fasting glucose rate-of-change (β = -0.05 mmol/L/year per allele, p = 0.058 among women; β = -0.004 mmol/L/year per allele, p = 0.863 among men), after accounting for baseline variables. Mean fasting glucose and mean BMI increased over an eight-year period and a substantial number of individuals developed type 2 diabetes by 2018. However, fasting glucose rate-of-change, and type 2 diabetes development was lower among females with AG and AA genotypes.

CONCLUSIONS

Further research is needed to understand the effect of the A allele on fasting glucose and type 2 diabetes development. Based on our observations that other risk factors increased over time, we advocate for the continued promotion for diabetes prevention and treatment programming, and the reduction of modifiable risk factors, in this setting.

Imputation accuracy across global human populations

Genotype imputation is now fundamental for genome-wide association studies but lacks fairness due to the underrepresentation of references from non-European ancestries. The state-of-the-art imputation reference panel released by the Trans-Omics for Precision Medicine (TOPMed) initiative improved the imputation of admixed African-ancestry and Hispanic/Latino samples, but imputation for populations primarily residing outside of North America may still fall short in performance due to persisting underrepresentation. To illustrate this point, we imputed the genotypes of over 43,000 individuals across 123 populations around the world and identified numerous populations where imputation accuracy paled in comparison to that of European-ancestry populations. For instance, the mean imputation r-squared (Rsq) for variants with minor allele frequencies between 1% and 5% in Saudi Arabians (n = 1,061), Vietnamese (n = 1,264), Thai (n = 2,435), and Papua New Guineans (n = 776) were 0.79, 0.78, 0.76, and 0.62, respectively, compared to 0.90-0.93 for comparable European populations matched in sample size and SNP array content. Outside of Africa and Latin America, Rsq appeared to decrease as genetic distances to European-ancestry reference increased, as predicted. Using sequencing data as ground truth, we also showed that Rsq may over-estimate imputation accuracy for non-European populations more than European populations, suggesting further disparity in accuracy between populations. Using 1,496 sequenced individuals from Taiwan Biobank as a second reference panel to TOPMed, we also assessed a strategy to improve imputation for non-European populations with meta-imputation, but this design did not improve accuracy across frequency spectra. Taken together, our analyses suggest that we must ultimately strive to increase diversity and size to promote equity within genetics research.

Recombination map tailored to Native Hawaiians may improve robustness of genomic scans for positive selection

Recombination events establish the patterns of haplotypic structure in a population and estimates of recombination rates are used in several downstream population and statistical genetic analyses. Using suboptimal maps from distantly related populations may reduce the efficacy of genomic analyses, particularly for underrepresented populations such as the Native Hawaiians. To overcome this challenge, we constructed recombination maps using genome-wide array data from two study samples of Native Hawaiians: one reflecting the current admixed state of Native Hawaiians (NH map) and one based on individuals of enriched Polynesian ancestries (PNS map) with the potential to be used for less admixed Polynesian populations such as the Samoans. We found the recombination landscape to be less correlated with those from other continental populations (e.g. Spearman's rho = 0.79 between PNS and CEU (Utah residents with Northern and Western European ancestry) compared to 0.92 between YRI (Yoruba in Ibadan, Nigeria) and CEU at 50 kb resolution), likely driven by the unique demographic history of the Native Hawaiians. PNS also shared the fewest recombination hotspots with other populations (e.g. 8% of hotspots shared between PNS and CEU compared to 27% of hotspots shared between YRI and CEU). We found that downstream analyses in the Native Hawaiian population, such as local ancestry inference, imputation, and IBD segment and relatedness detections, would achieve similar efficacy when using the NH map compared to an omnibus map. However, for genome scans of adaptive loci using integrated haplotype scores, we found several loci with apparent genome-wide significant signals (|Z-score|> 4) in Native Hawaiians that would not have been significant when analyzed using NH-specific maps. Population-specific recombination maps may therefore improve the robustness of haplotype-based statistics and help us better characterize the evolutionary history that may underlie Native Hawaiian-specific health conditions that persist today.

The accuracy of polygenic score models for anthropometric traits and Type II Diabetes in the Native Hawaiian Population

Polygenic scores (PGS) are promising in stratifying individuals based on the genetic susceptibility to complex diseases or traits. However, the accuracy of PGS models, typically trained in European- or East Asian-ancestry populations, tend to perform poorly in other ethnic minority populations, and their accuracies have not been evaluated for Native Hawaiians. Using body mass index, height, and type-2 diabetes as examples of highly polygenic traits, we evaluated the prediction accuracies of PGS models in a large Native Hawaiian sample from the Multiethnic Cohort with up to 5,300 individuals. We evaluated both publicly available PGS models or genome-wide PGS models trained in this study using the largest available GWAS. We found evidence of lowered prediction accuracies for the PGS models in some cases, particularly for height. We also found that using the Native Hawaiian samples as an optimization cohort during training did not consistently improve PGS performance. Moreover, even the best performing PGS models among Native Hawaiians would have lowered prediction accuracy among the subset of individuals most enriched with Polynesian ancestry. Our findings indicate that factors such as admixture histories, sample size and diversity in GWAS can influence PGS performance for complex traits among Native Hawaiian samples. This study provides an initial survey of PGS performance among Native Hawaiians and exposes the current gaps and challenges associated with improving polygenic prediction models for underrepresented minority populations.

Tree-based QTL mapping with expected local genetic relatedness matrices

Understanding the genetic basis of complex phenotypes is a central pursuit of genetics. Genome-wide association studies (GWASs) are a powerful way to find genetic loci associated with phenotypes. GWASs are widely and successfully used, but they face challenges related to the fact that variants are tested for association with a phenotype independently, whereas in reality variants at different sites are correlated because of their shared evolutionary history. One way to model this shared history is through the ancestral recombination graph (ARG), which encodes a series of local coalescent trees. Recent computational and methodological breakthroughs have made it feasible to estimate approximate ARGs from large-scale samples. Here, we explore the potential of an ARG-based approach to quantitative-trait locus (QTL) mapping, echoing existing variance-components approaches. We propose a framework that relies on the conditional expectation of a local genetic relatedness matrix (local eGRM) given the ARG. Simulations show that our method is especially beneficial for finding QTLs in the presence of allelic heterogeneity. By framing QTL mapping in terms of the estimated ARG, we can also facilitate the detection of QTLs in understudied populations. We use local eGRM to analyze two chromosomes containing known body size loci in a sample of Native Hawaiians. Our investigations can provide intuition about the benefits of using estimated ARGs in population- and statistical-genetic methods in general.

Digital Determinants of Health: Health data poverty amplifies existing health disparities-A scoping review

Artificial intelligence (AI) and machine learning (ML) have an immense potential to transform healthcare as already demonstrated in various medical specialties. This scoping review focuses on the factors that influence health data poverty, by conducting a literature review, analysis, and appraisal of results. Health data poverty is often an unseen factor which leads to perpetuating or exacerbating health disparities. Improvements or failures in addressing health data poverty will directly impact the effectiveness of AI/ML systems. The potential causes are complex and may enter anywhere along the development process. The initial results highlighted studies with common themes of health disparities (72%), AL/ML bias (28%) and biases in input data (18%). To properly evaluate disparities that exist we recommend a strengthened effort to generate unbiased equitable data, improved understanding of the limitations of AI/ML tools, and rigorous regulation with continuous monitoring of the clinical outcomes of deployed tools.

Recombination map tailored to Native Hawaiians improves robustness of genomic scans for positive selection

Recombination events establish the patterns of haplotypic structure in a population and estimates of recombination rates are used in several downstream population and statistical genetic analyses. Using suboptimal maps from distantly related populations may reduce the efficacy of genomic analyses, particularly for underrepresented populations such as the Native Hawaiians. To overcome this challenge, we constructed recombination maps using genome-wide array data from two study samples of Native Hawaiians: one reflecting the current admixed state of Native Hawaiians (NH map), and one based on individuals of enriched Polynesian ancestries (PNS map) with the potential to be used for less admixed Polynesian populations such as the Samoans. We found the recombination landscape to be less correlated with those from other continental populations (e.g. Spearman's rho = 0.79 between PNS and CEU (Utah residents with Northern and Western European ancestry) compared to 0.92 between YRI (Yoruba in Ibadan, Nigeria) and CEU at 50 kb resolution), likely driven by the unique demographic history of the Native Hawaiians. PNS also shared the fewest recombination hotspots with other populations (e.g. 8% of hotspots shared between PNS and CEU compared to 27% of hotspots shared between YRI and CEU). We found that downstream analyses in the Native Hawaiian population, such as local ancestry inference, imputation, and IBD segment and relatedness detections, would achieve similar efficacy when using the NH map compared to an omnibus map. However, for genome scans of adaptive loci using integrated haplotype scores, we found several loci with apparent genome-wide significant signals (|Z-score| > 4) in Native Hawaiians that would not have been significant when analyzed using NH-specific maps. Population-specific recombination maps may therefore improve the robustness of haplotype-based statistics and help us better characterize the evolutionary history that may underlie Native Hawaiian-specific health conditions that persist today.

Tree-based QTL mapping with expected local genetic relatedness matrices

Understanding the genetic basis of complex phenotypes is a central pursuit of genetics. Genome-wide Association Studies (GWAS) are a powerful way to find genetic loci associated with phenotypes. GWAS are widely and successfully used, but they face challenges related to the fact that variants are tested for association with a phenotype independently, whereas in reality variants at different sites are correlated because of their shared evolutionary history. One way to model this shared history is through the ancestral recombination graph (ARG), which encodes a series of local coalescent trees. Recent computational and methodological breakthroughs have made it feasible to estimate approximate ARGs from large-scale samples. Here, we explore the potential of an ARG-based approach to quantitative-trait locus (QTL) mapping, echoing existing variance-components approaches. We propose a framework that relies on the conditional expectation of a local genetic relatedness matrix given the ARG (local eGRM). Simulations show that our method is especially beneficial for finding QTLs in the presence of allelic heterogeneity. By framing QTL mapping in terms of the estimated ARG, we can also facilitate the detection of QTLs in understudied populations. We use local eGRM to identify a large-effect BMI locus, the CREBRF gene, in a sample of Native Hawaiians in which it was not previously detectable by GWAS because of a lack of population-specific imputation resources. Our investigations can provide intuition about the benefits of using estimated ARGs in population- and statistical-genetic methods in general.

Multivariate analysis of a missense variant in CREBRF reveals associations with measures of adiposity in people of Polynesian ancestries

The minor allele of rs373863828, a missense variant in CREB3 Regulatory Factor, is associated with several cardiometabolic phenotypes in Polynesian peoples. To better understand the variant, we tested the association of rs373863828 with a panel of correlated phenotypes (body mass index [BMI], weight, height, HDL cholesterol, triglycerides, and total cholesterol) using multivariate Bayesian association and network analyses in a Samoa cohort (n = 1632), Aotearoa New Zealand cohort (n = 1419), and combined cohort (n = 2976). An expanded set of phenotypes (adding estimated fat and fat-free mass, abdominal circumference, hip circumference, and abdominal-hip ratio) was tested in the Samoa cohort (n = 1496). In the Samoa cohort, we observed significant associations (log10 Bayes Factor [BF] ≥ 5.0) between rs373863828 and the overall phenotype panel (8.81), weight (8.30), and BMI (6.42). In the Aotearoa New Zealand cohort, we observed suggestive associations (1.5 < log10 BF < 5) between rs373863828 and the overall phenotype panel (4.60), weight (3.27), and BMI (1.80). In the combined cohort, we observed concordant signals with larger log10 BFs. In the Samoa-specific expanded phenotype analyses, we also observed significant associations between rs373863828 and fat mass (5.65), abdominal circumference (5.34), and hip circumference (5.09). Bayesian networks provided evidence for a direct association of rs373863828 with weight and indirect associations with height and BMI.

Genetic alterations in CREBRF influence prostate cancer survival and impact prostate tissue homeostasis in mice

BACKGROUND

Risk factors for prostate cancer include age, environment, race and ethnicity. Genetic variants in cyclic-adenosine-monophosphate-response-element-binding protein 3 regulatory factor (CREBRF) gene are frequently observed in Pacific Islanders, a population with elevated prostate cancer incidence. CREBRF has been shown to play a role in other cancers, however its function in prostate homeostasis and tumorigenesis has not been previously explored. We determined the incidence of CREBRF alterations in publicly available databases and examined the impact of CREBRF deletion on the murine prostate in order to determine whether CREBRF impacts prostate physiology or pathophysiology.

METHODS

Alterations in CREBRF were identified in prostate cancer patients via in silico analysis of several publicly available datasets through cBioPortal. Male Crebrf knockout and wild-type littermate mice were generated and examined for prostate defects at 4 months of age. Immunohistochemical staining of murine prostate sections was used to determine the impact of Crebrf knockout on proliferation, apoptosis, inflammation and blood vessel density in the prostate. Serum adipokine levels were measured using a Luminex Multiplex Assay.

RESULTS

CREBRF alterations were identified in up to 4.05% of prostate tumors and the mutations identified were categorized as likely damaging. Median survival of prostate cancer patients with genetic alterations in CREBRF was 41.23 months, compared to 131 months for patients without these changes. In the murine model, the prostates of Crebrf knockout mice had reduced epithelial proliferation and increased TUNEL⁺ apoptotic cells. Circulating adipokines PAI-1 and MCP-1 were also altered in Crebrf knockout mice compared to age-matched controls.

CONCLUSIONS

Prostate cancer patients with genetic alterations in CREBRF had a significantly decreased overall survival suggesting that wild type CREBRF may play a role in limiting prostate tumorigenesis and progression. The murine knockout model demonstrated that CREBRF could modulate proliferation and apoptosis and macrophage density in the prostate. Serum levels of adipokines PAI-1 and MCP-1 were also altered and may contribute to the phenotypic changes observed in the prostates of Crebrf knockout mice. Future studies focused on populations susceptible to CREBRF mutations and mechanistic studies will be required to fully elucidate the potential role of CREBRF in prostate tumorigenesis.

Developing CIRdb as a catalog of natural genetic variation in the Canary Islanders

The current inhabitants of the Canary Islands have a unique genetic makeup in the European diversity landscape due to the existence of African footprints from recent admixture events, especially of North African components (> 20%). The underrepresentation of non-Europeans in genetic studies and the sizable North African ancestry, which is nearly absent from all existing catalogs of worldwide genetic diversity, justify the need to develop CIRdb, a population-specific reference catalog of natural genetic variation in the Canary Islanders. Based on array genotyping of the selected unrelated donors and comparisons against available datasets from European, sub-Saharan, and North African populations, we illustrate the intermediate genetic differentiation of Canary Islanders between Europeans and North Africans and the existence of within-population differences that are likely driven by genetic isolation. Here we describe the overall design and the methods that are being implemented to further develop CIRdb. This resource will help to strengthen the implementation of Precision Medicine in this population by contributing to increase the diversity in genetic studies. Among others, this will translate into improved ability to fine map disease genes and simplify the identification of causal variants and estimate the prevalence of unattended Mendelian diseases.

A genealogical estimate of genetic relationships

The application of genetic relationships among individuals, characterized by a genetic relationship matrix (GRM), has far-reaching effects in human genetics. However, the current standard to calculate the GRM treats linked markers as independent and does not explicitly model the underlying genealogical history of the study sample. Here, we propose a coalescent-informed framework, namely the expected GRM (eGRM), to infer the expected relatedness between pairs of individuals given an ancestral recombination graph (ARG) of the sample. Through extensive simulations, we show that the eGRM is an unbiased estimate of latent pairwise genome-wide relatedness and is robust when computed with ARG inferred from incomplete genetic data. As a result, the eGRM better captures the structure of a population than the canonical GRM, even when using the same genetic information. More importantly, our framework allows a principled approach to estimate the eGRM at different time depths of the ARG, thereby revealing the time-varying nature of population structure in a sample. When applied to SNP array genotypes from a population sample from Northern and Eastern Finland, we find that clustering analysis with the eGRM reveals population structure driven by subpopulations that would not be apparent via the canonical GRM and that temporally the population model is consistent with recent divergence and expansion. Taken together, our proposed eGRM provides a robust tree-centric estimate of relatedness with wide application to genetic studies.

Limited Metabolic Effect of the CREBRFR457Q Obesity Variant in Mice

The Arg457Gln missense variant in the CREBRF gene has previously been identified as driving excess body weight in Pacific/Oceanic populations. Intriguingly, Arg457Gln variant carriers also demonstrate paradoxical reductions in diabetes risk, indicating that the gene has a critical role in whole-body metabolism. To study the function of this variant in more detail, we generated mice on an FVB/N background with the Crebrf Arg458Gln variant knocked in to replace the endogenous Crebrf. The whole-body metabolic phenotype was characterized for male and female mice on a regular chow diet or an 8-week high-fat challenge. Regular assessment of body composition found that the Crebrf variant had no influence on total body weight or fat mass at any time point. Glucose tolerance tests demonstrated no obvious genotype effect on glucose homeostasis, with indirect calorimetry measures of whole-body energy expenditure likewise unaffected. Male chow-fed variant carriers displayed a trend towards increased lean mass and significantly reduced sensitivity to insulin administration. Overall, this novel mouse model showed only limited phenotypic effects associated with the Crebrf missense variant. The inability to recapitulate results of human association studies may invite reconsideration of the precise mechanistic link between CREBRF function and the risks of obesity and diabetes in variant allele carriers.

Using population-specific add-on polymorphisms to improve genotype imputation in underrepresented populations

Genome-wide association studies rely on the statistical inference of untyped variants, called imputation, to increase the coverage of genotyping arrays. However, the results are often suboptimal in populations underrepresented in existing reference panels and array designs, since the selected single nucleotide polymorphisms (SNPs) may fail to capture population-specific haplotype structures, hence the full extent of common genetic variation. Here, we propose to sequence the full genomes of a small subset of an underrepresented study cohort to inform the selection of population-specific add-on tag SNPs and to generate an internal population-specific imputation reference panel, such that the remaining array-genotyped cohort could be more accurately imputed. Using a Tanzania-based cohort as a proof-of-concept, we demonstrate the validity of our approach by showing improvements in imputation accuracy after the addition of our designed add-on tags to the base H3Africa array.

Genome-wide association studies, which study the association between genetic variants and various phenotypes, typically rely on genotyping arrays. Only a small proportion of genetic variants within the genome are typed on genotyping arrays. Untyped variants are statistically inferred through a process known as genotype imputation, where correlations between variants (haplotypes) observed in external reference panels are leveraged to infer untyped variants in the study population. However, for study populations that are underrepresented in existing reference panels, the quality of imputation is often sub-optimal. This is because typed variants incorporated on existing genotyping arrays can be unsuitable for the study population, and haplotype structures can be different between the reference and the study population. Here, we illustrate an approach to select a custom set of population-specific typed variants to improve genotype imputation in such underrepresented populations.

The CREBRF diabetes-protective rs373863828-A allele is associated with enhanced early insulin release in men of Māori and Pacific ancestry

AIMS/HYPOTHESIS

The minor A allele of rs373863828 (CREBRF p.Arg457Gln) is associated with increased BMI, but reduced risk of type 2 and gestational diabetes in Polynesian (Pacific peoples and Aotearoa New Zealand Māori) populations. This study investigates the effect of the A allele on insulin release and sensitivity in overweight/obese men without diabetes.

METHODS

A mixed meal tolerance test was completed by 172 men (56 with the A allele) of Māori or Pacific ancestry, and 44 (24 with the A allele) had a frequently sampled IVGTT and hyperinsulinaemic-euglycaemic clamp. Mixed linear models with covariates age, ancestry and BMI were used to analyse the association between the A allele of rs373863828 and markers of insulin release and blood glucose regulation.

RESULTS

The A allele of rs373863828 is associated with a greater increase in plasma insulin 30 min following a meal challenge without affecting the elevation in plasma glucose or incretins glucagon-like polypeptide-1 or gastric inhibitory polypeptide. Consistent with this point, following an i.v. infusion of a glucose bolus, participants with an A allele had higher early (p < 0.05 at 2 and 4 min) plasma insulin and C-peptide concentrations for a similar elevation in blood glucose as those homozygous for the major (G) allele. Despite increased plasma insulin, rs373863828 genotype was not associated with a significant difference (p > 0.05) in insulin sensitivity index or glucose disposal during hyperinsulinaemic-euglycaemic clamp.

CONCLUSIONS/INTERPRETATION

rs373863828-A allele associates with increased glucose-stimulated insulin release without affecting insulin sensitivity, suggesting that CREBRF p.Arg457Gln may increase insulin release to reduce the risk of type 2 diabetes.

The Opportunities and Challenges of Integrating Population Histories Into Genetic Studies for Diverse Populations: A Motivating Example From Native Hawaiians

There is a well-recognized need to include diverse populations in genetic studies, but several obstacles continue to be prohibitive, including (but are not limited to) the difficulty of recruiting individuals from diverse populations in large numbers and the lack of representation in available genomic references. These obstacles notwithstanding, studying multiple diverse populations would provide informative, population-specific insights. Using Native Hawaiians as an example of an understudied population with a unique evolutionary history, I will argue that by developing key genomic resources and integrating evolutionary thinking into genetic epidemiology, we will have the opportunity to efficiently advance our knowledge of the genetic risk factors, ameliorate health disparity, and improve healthcare in this underserved population.

A murine model of the human CREBRFR457Q obesity-risk variant does not influence energy or glucose homeostasis in response to nutritional stress

Obesity and diabetes have strong heritable components, yet the genetic contributions to these diseases remain largely unexplained. In humans, a missense variant in Creb3 regulatory factor (CREBRF) [rs373863828 (p.Arg457Gln); CREBRFR457Q] is strongly associated with increased odds of obesity but decreased odds of diabetes. Although virtually nothing is known about CREBRF's mechanism of action, emerging evidence implicates it in the adaptive transcriptional response to nutritional stress downstream of TORC1. The objectives of this study were to generate a murine model with knockin of the orthologous variant in mice (CREBRFR458Q) and to test the hypothesis that this CREBRF variant promotes obesity and protects against diabetes by regulating energy and glucose homeostasis downstream of TORC1. To test this hypothesis, we performed extensive phenotypic analysis of CREBRFR458Q knockin mice at baseline and in response to acute (fasting/refeeding), chronic (low- and high-fat diet feeding), and extreme (prolonged fasting) nutritional stress as well as with pharmacological TORC1 inhibition, and aging to 52 weeks. The results demonstrate that the murine CREBRFR458Q model of the human CREBRFR457Q variant does not influence energy/glucose homeostasis in response to these interventions, with the exception of possible greater loss of fat relative to lean mass with age. Alternative preclinical models and/or studies in humans will be required to decipher the mechanisms linking this variant to human health and disease.

A multi-phenotype genome-wide association study of clades causing tuberculosis in a Ghanaian- and South African cohort

Despite decades of research and advancements in diagnostics and treatment, tuberculosis remains a major public health concern. New computational methods are needed to interrogate the intersection of host- and bacterial genomes. Paired host genotype datum and infecting bacterial isolate information were analysed for associations using a multinomial logistic regression framework implemented in SNPTest. A cohort of 853 admixed South African participants and a Ghanaian cohort of 1359 participants were included. Two directly genotyped variants, namely rs529920 and rs41472447, were identified in the Ghanaian cohort as being statistically significantly associated with risk for infection with strains of different members of the MTBC. Thus, a multinomial logistic regression using paired host-pathogen data may prove valuable for investigating the complex relationships driving infectious disease.

The impact of global and local Polynesian genetic ancestry on complex traits in Native Hawaiians

Epidemiological studies of obesity, Type-2 diabetes (T2D), cardiovascular diseases and several common cancers have revealed an increased risk in Native Hawaiians compared to European- or Asian-Americans living in the Hawaiian islands. However, there remains a gap in our understanding of the genetic factors that affect the health of Native Hawaiians. To fill this gap, we studied the genetic risk factors at both the chromosomal and sub-chromosomal scales using genome-wide SNP array data on ~4,000 Native Hawaiians from the Multiethnic Cohort. We estimated the genomic proportion of Native Hawaiian ancestry ("global ancestry," which we presumed to be Polynesian in origin), as well as this ancestral component along each chromosome ("local ancestry") and tested their respective association with binary and quantitative cardiometabolic traits. After attempting to adjust for non-genetic covariates evaluated through questionnaires, we found that per 10% increase in global Polynesian genetic ancestry, there is a respective 8.6%, and 11.0% increase in the odds of being diabetic (P = 1.65×10-4) and having heart failure (P = 2.18×10-4), as well as a 0.059 s.d. increase in BMI (P = 1.04×10-10). When testing the association of local Polynesian ancestry with risk of disease or biomarkers, we identified a chr6 region associated with T2D. This association was driven by an uniquely prevalent variant in Polynesian ancestry individuals. However, we could not replicate this finding in an independent Polynesian cohort from Samoa due to the small sample size of the replication cohort. In conclusion, we showed that Polynesian ancestry, which likely capture both genetic and lifestyle risk factors, is associated with an increased risk of obesity, Type-2 diabetes, and heart failure, and that larger cohorts of Polynesian ancestry individuals will be needed to replicate the putative association on chr6 with T2D.

Tractor uses local ancestry to enable the inclusion of admixed individuals in GWAS and to boost power

Admixed populations are routinely excluded from genomic studies due to concerns over population structure. Here, we present a statistical framework and software package, Tractor, to facilitate the inclusion of admixed individuals in association studies by leveraging local ancestry. We test Tractor with simulated and empirical two-way admixed African-European cohorts. Tractor generates accurate ancestry-specific effect-size estimates and P values, can boost genome-wide association study (GWAS) power and improves the resolution of association signals. Using a local ancestry-aware regression model, we replicate known hits for blood lipids, discover novel hits missed by standard GWAS and localize signals closer to putative causal variants.