Precision arbovirus serology with a pan-arbovirus peptidome.

Arthropod-borne viruses represent a crucial public health threat. Current arboviral serology assays are either labor intensive or incapable of distinguishing closely related viruses, and many zoonotic arboviruses that may transition to humans lack any serologic assays. In this study, we present a programmable phage display platform, ArboScan, that evaluates antibody binding to overlapping peptides that represent the proteomes of 691 human and zoonotic arboviruses. We confirm that ArboScan provides detailed antibody binding information from animal sera, human sera, and an arthropod blood meal. ArboScan identifies distinguishing features of antibody responses based on exposure history in a Colombian cohort of Zika patients. Finally, ArboScan details epitope level information that rapidly identifies candidate epitopes with potential protective significance. ArboScan thus represents a resource for characterizing human and animal arbovirus antibody responses at cohort scale.

Custom Biomedical FAIR Data Analysis in the Cloud Using CAVATICA.

The historically fragmented biomedical data ecosystem has moved towards harmonization under the findable, accessible, interoperable, and reusable (FAIR) data principles, creating more opportunities for cloud-based research. This shift is especially opportune for scientists across diverse domains interested in implementing creative, nonstandard computational analytic pipelines on large and varied datasets. However, executing custom cloud analyses may present difficulties, particularly for investigators lacking advanced computational expertise. Here, we present an accessible, streamlined approach for the cloud compute platform CAVATICA that offers a solution. We outline how we developed a custom workflow in the cloud, for analyzing whole genome sequences of case-parent trios to detect sex-specific genetic effects on orofacial cleft risk, which required several programming languages and custom software packages. The approach involves just three components: Docker to containerize software environments, tool creation for each analysis step, and a visual workflow editor to weave the tools into a Common Workflow Language (CWL) pipeline. Our approach should be accessible to any investigator with basic computational skills, is readily extended to implement any scalable high-throughput biomedical data analysis in the cloud, and is applicable to other commonly used compute platforms such as BioData Catalyst. We believe our approach empowers versatile data reuse and promotes accelerated biomedical discovery in a time of substantial FAIR data.

Multi-omics in nasal epithelium reveals three axes of dysregulation for asthma risk in the African Diaspora populations.

Asthma has striking disparities across ancestral groups, but the molecular underpinning of these differences is poorly understood and minimally studied. A goal of the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA) is to understand multi-omic signatures of asthma focusing on populations of African ancestry. RNASeq and DNA methylation data are generated from nasal epithelium including cases (current asthma, N = 253) and controls (never-asthma, N = 283) from 7 different geographic sites to identify differentially expressed genes (DEGs) and gene networks. We identify 389 DEGs; the top DEG, FN1, was downregulated in cases (q = 3.26 × 10-9) and encodes fibronectin which plays a role in wound healing. The top three gene expression modules implicate networks related to immune response (CEACAM5; p = 9.62 × 10-16 and CPA3; p = 2.39 × 10-14) and wound healing (FN1; p = 7.63 × 10-9). Multi-omic analysis identifies FKBP5, a co-chaperone of glucocorticoid receptor signaling known to be involved in drug response in asthma, where the association between nasal epithelium gene expression is likely regulated by methylation and is associated with increased use of inhaled corticosteroids. This work reveals molecular dysregulation on three axes - increased Th2 inflammation, decreased capacity for wound healing, and impaired drug response - that may play a critical role in asthma within the African Diaspora.

Efficient encoding of large antigenic spaces by epitope prioritization with Dolphyn.

We investigate a relatively underexplored component of the gut-immune axis by profiling the antibody response to gut phages using Phage Immunoprecipitation Sequencing (PhIP-Seq). To cover large antigenic spaces, we develop Dolphyn, a method that uses machine learning to select peptides from protein sets and compresses the proteome through epitope-stitching. Dolphyn compresses the size of a peptide library by 78% compared to traditional tiling, increasing the antibody-reactive peptides from 10% to 31%. We find that the immune system develops antibodies to human gut bacteria-infecting viruses, particularly E.coli-infecting Myoviridae. Cost-effective PhIP-Seq libraries designed with Dolphyn enable the assessment of a wider range of proteins in a single experiment, thus facilitating the study of the gut-immune axis.

Comprehensive profiling of pre-infection antibodies identifies HIV targets associated with viremic control and viral load.

Background: High HIV viral load (VL) is associated with increased transmission risk and faster disease progression. HIV controllers achieve viral suppression without antiretroviral (ARV) treatment. We evaluated viremic control in a community-randomized trial with >48,000 participants. Methods: A massively multiplexed antibody profiling system, VirScan, was used to quantify pre- and post-infection antibody reactivity to HIV peptides in 664 samples from 429 participants (13 controllers, 135 viremic non-controllers, 64 other non-controllers, 217 uninfected persons). Controllers had VLs <2,000 copies/mL with no ARV drugs detected at the first HIV-positive visit and one year later. Viremic non-controllers had VLs 2,000 copies/mL with no ARV drugs detected at the first HIV-positive visit. Other non-controllers had either ARV drugs detected at the first HIV-positive visit (n=47) or VLs <2,000 copies/mL with no ARV drugs detected at only one HIV-positive visit (n=17). Results: We identified pre-infection HIV antibody reactivities that correlated with post-infection VL. Pre-infection reactivity to an epitope in the HR2 domain of gp41 was associated with controller status and lower VL. Pre-infection reactivity to an epitope in the C2 domain of gp120 was associated with non-controller status and higher VL. Different patterns of antibody reactivity were observed over time for these two epitopes. Conclusion: These studies suggest that pre-infection HIV antibodies are associated with controller status and modulation of HIV VL. These findings may inform research on antibody-based interventions for HIV treatment.

Efficient encoding of large antigenic spaces by epitope prioritization with Dolphyn.

We investigated a relatively underexplored component of the gut-immune axis by profiling the antibody response to gut phages using Phage Immunoprecipitation Sequencing (PhIP-Seq). To enhance this approach, we developed Dolphyn, a novel method that uses machine learning to select peptides from protein sets and compresses the proteome through epitope-stitching. Dolphyn improves the fraction of gut phage library peptides bound by antibodies from 10% to 31% in healthy individuals, while also reducing the number of synthesized peptides by 78%. In our study on gut phages, we discovered that the immune system develops antibodies to bacteria-infecting viruses in the human gut, particularly E.coli-infecting Myoviridae. Cost-effective PhIP-Seq libraries designed with Dolphyn enable the assessment of a wider range of proteins in a single experiment, thus facilitating the study of the gut-immune axis.

Novel bioinformatic methods and machine learning approaches reveal candidate biomarkers of the intensity and timing of past exposure to Plasmodium falciparum.

Accurately quantifying the burden of malaria over time is an important goal of malaria surveillance efforts and can enable effective targeting and evaluation of interventions. Malaria surveillance methods capture active or recent infections which poses several challenges to achieving malaria surveillance goals. In high transmission settings, asymptomatic infections are common and therefore accurate measurement of malaria burden demands active surveillance; in low transmission regions where infections are rare accurate surveillance requires sampling large subsets of the population; and in any context monitoring malaria burden over time necessitates serial sampling. Antibody responses to Plasmodium falciparum parasites persist after infection and therefore measuring antibodies has the potential to overcome several of the current obstacles to accurate malaria surveillance. Identifying which antibody responses are markers of the timing and intensity of past exposure to P. falciparum remains challenging, particularly among adults who tend to be re-exposed multiple times over the course of their lifetime and therefore have similarly high antibody responses to many Plasmodium antigens. A previous analysis of 479 serum samples from individuals in three regions in southern Africa with different historical levels of P. falciparum malaria transmission (high, intermediate, and low) revealed regional differences in antibody responses to P. falciparum antigens among children under 5 years of age. Using a novel bioinformatic pipeline optimized for protein microarrays that minimizes between-sample technical variation, we used antibody responses to Plasmodium antigens as predictors in random forest models to classify samples from adults into these three regions of differing historical malaria transmission with high accuracy (AUC = 0.99). Many of the most important antigens for classification in these models do not overlap with previously published results and are therefore novel candidate markers for the timing and intensity of past exposure to P. falciparum. Measuring antibody responses to these antigens could lead to improved malaria surveillance.

A Bayesian hierarchical model for signal extraction from protein microarrays.

Protein microarrays are a promising technology that measure protein levels in serum or plasma samples. Due to their high technical variability and high variation in protein levels across serum samples in any population, directly answering biological questions of interest using protein microarray measurements is challenging. Analyzing preprocessed data and within-sample ranks of protein levels can mitigate the impact of between-sample variation. As for any analysis, ranks are sensitive to preprocessing, but loss function based ranks that accommodate major structural relations and components of uncertainty are very effective. Bayesian modeling with full posterior distributions for quantities of interest produce the most effective ranks. Such Bayesian models have been developed for other assays, for example, DNA microarrays, but modeling assumptions for these assays are not appropriate for protein microarrays. Consequently, we develop and evaluate a Bayesian model to extract the full posterior distribution of normalized protein levels and associated ranks for protein microarrays, and show that it fits well to data from two studies that use protein microarrays produced by different manufacturing processes. We validate the model via simulation and demonstrate the downstream impact of using estimates from this model to obtain optimal ranks.

Epidermal differentiation complex genetic variation in atopic dermatitis and peanut allergy.

BACKGROUND: Deleterious variation in the epidermal differentiation complex (EDC) on chromosome 1 is a well-known genetic determinant of atopic dermatitis (AD) and has been associated with risk of peanut allergy (PA) in population-based studies. OBJECTIVE: Our aim was to determine the effect of genetic variation in the EDC on AD trajectory and risk of PA in early life. METHODS: Genome sequencing was used to measure genetic variation in the EDC in the Learning Early about Peanut Allergy (LEAP) study participants. Association tests were done to identify gene- and variant-level predicted deleterious variation associated with AD severity by using the Scoring Atopic Dermatitis (SCORAD) tool (n = 559) at baseline and each follow-up visit, as well as PA and food allergy in peanut avoiders (n = 275). Predicted deleterious variants included missense variants that were frameshift insertions, frameshift deletions, stop-gain mutations, or stop-loss mutations. Associations between variant load, SCORAD score, and PA were tested by using linear and generalized linear regression models. RESULTS: The genes FLG, FLG2, HRNR, and TCHH1 harbored the most predicted deleterious variation (30, 6, 3, and 1 variant, respectively). FLG variants were associated with SCORAD score at all time points; 4 variants (R1798X, R501X, S126X, and S761fs) drove the association with SCORAD score at each time point, and higher variant load was associated with greater AD severity over time. There was an association between these variants and PA, which remained significant independent of baseline AD severity (odds ratio = 2.63 [95% CI = 1.11-6.01] [P = .02]). CONCLUSIONS: Variation in FLG predicted to be deleterious is associated with AD severity at baseline and longitudinally and has an association with PA independent of baseline severity.

HLA-associated outcomes in peanut oral immunotherapy trials identify mechanistic and clinical determinants of therapeutic success.

Rationale: Previous studies identified an interaction between HLA and oral peanut exposure. HLA-DQA1*01:02 had a protective role with the induction of Ara h 2 epitope-specific IgG4 associated with peanut consumption during the LEAP clinical trial for prevention of peanut allergy, while it was a risk allele for peanut allergy in the peanut avoidance group. We have now evaluated this gene-environment interaction in two subsequent peanut oral immunotherapy (OIT) trials - IMPACT and POISED - to better understand the potential for the HLA-DQA1*01:02 allele as an indicator of higher likelihood of desensitization, sustained unresponsiveness, and peanut allergy remission. Methods: We determined HLA-DQA1*01:02 carrier status using genome sequencing from POISED (N=118, age: 7-55yr) and IMPACT (N=126, age: 12-<48mo). We tested for association with remission, sustained unresponsiveness (SU), and desensitization in the OIT groups, as well as peanut component specific IgG4 (psIgG4) using generalized linear models and adjusting for relevant covariates and ancestry. Results: While not quite statistically significant, a higher proportion of HLA-DQA1*01:02 carriers receiving OIT in IMPACT were desensitized (93%) compared to non-carriers (78%); odds ratio (OR)=5.74 (p=0.06). In this sample we also observed that a higher proportion of carriers achieved remission (35%) compared to non-carriers (22%); OR=1.26 (p=0.80). In POISED, carriers more frequently attained continued desensitization (80% versus 61% among non-carriers; OR=1.28, p=0.86) and achieved SU (52% versus 31%; OR=2.32, p=0.19). psIgG4 associations with HLA-DQA1*01:02 in the OIT arm of IMPACT which included younger study subjects recapitulated patterns noted in LEAP, but no associations of note were observed in the older POISED study subjects. Conclusions: Findings across three clinical trials show a pattern of a gene environment interaction between HLA and oral peanut exposure. Age, and prior sensitization contribute additional determinants of outcomes, consistent with a mechanism of restricted antigen recognition fundamental to driving protective immune responses to OIT.

Evaluation of multi-assay algorithms for cross-sectional HIV incidence estimation in settings with universal antiretroviral treatment.

BACKGROUND: Multi-assay algorithms (MAAs) are used to estimate population-level HIV incidence and identify individuals with recent infection. Many MAAs use low viral load (VL) as a biomarker for long-term infection. This could impact incidence estimates in settings with high rates of early HIV treatment initiation. We evaluated the performance of two MAAs that do not include VL. METHODS: Samples were collected from 219 seroconverters (infected < 1 year) and 4376 non-seroconverters (infected > 1 year) in the HPTN 071 (PopART) trial; 28.8% of seroconverter samples and 73.2% of non-seroconverter samples had VLs ≤ 400 copies/mL. Samples were tested with the Limiting Antigen Avidity assay (LAg) and JHU BioRad-Avidity assays. Antibody reactivity to two HIV peptides was measured using the MSD U-PLEX assay. Two MAAs were evaluated that do not include VL: a MAA that includes the LAg-Avidity assay and BioRad-Avidity assay (LAg + BR) and a MAA that includes the LAg-Avidity assay and two peptide biomarkers (LAg + PepPair). Performance of these MAAs was compared to a widely used MAA that includes LAg and VL (LAg + VL). RESULTS: The incidence estimate for LAg + VL (1.29%, 95% CI: 0.97-1.62) was close to the observed longitudinal incidence (1.34% 95% CI: 1.17-1.53). The incidence estimates for the other two MAAs were higher (LAg + BR: 2.56%, 95% CI 2.01-3.11; LAg + PepPair: 2.84%, 95% CI: 1.36-4.32). LAg + BR and LAg + PepPair also misclassified more individuals infected > 2 years as recently infected than LAg + VL (1.2% [42/3483 and 1.5% [51/3483], respectively, vs. 0.2% [6/3483]). LAg + BR classified more seroconverters as recently infected than LAg + VL or LAg + PepPair (80 vs. 58 and 50, respectively) and identified ~ 25% of virally suppressed seroconverters as recently infected. CONCLUSIONS: The LAg + VL MAA produced a cross-sectional incidence estimate that was closer to the longitudinal estimate than two MAAs that did not include VL. The LAg + BR MAA classified the greatest number of individual seroconverters as recently infected but had a higher false recent rate.

Detecting antibody reactivities in Phage ImmunoPrecipitation Sequencing data.

Phage ImmunoPrecipitation Sequencing (PhIP-Seq) is a recently developed technology to assess antibody reactivity, quantifying antibody binding towards hundreds of thousands of candidate epitopes. The output from PhIP-Seq experiments are read count matrices, similar to RNA-Seq data; however some important differences do exist. In this manuscript we investigated whether the publicly available method edgeR (Robinson et al., Bioinformatics 26(1):139-140, 2010) for normalization and analysis of RNA-Seq data is also suitable for PhIP-Seq data. We find that edgeR is remarkably effective, but improvements can be made and introduce a Bayesian framework specifically tailored for data from PhIP-Seq experiments (Bayesian Enrichment Estimation in R, BEER).

Detecting and quantifying antibody reactivity in PhIP-Seq data with BEER.

SUMMARY: Because of their high abundance, easy accessibility in peripheral blood, and relative stability ex vivo, antibodies serve as excellent records of environmental exposures and immune responses. Phage Immuno-Precipitation Sequencing (PhIP-Seq) is the most efficient technique available for assessing antibody binding to hundreds of thousands of peptides at a cohort scale. PhIP-Seq is a high-throughput approach for assessing antibody reactivity to hundreds of thousands of candidate epitopes. Accurate detection of weakly reactive peptides is particularly important for characterizing the development and decline of antibody responses. Here, we present BEER (Bayesian Enrichment Estimation in R), a software package specifically developed for the quantification of peptide reactivity from PhIP-Seq experiments. BEER implements a hierarchical model and produces posterior probabilities for peptide reactivity and a fold change estimate to quantify the magnitude. BEER also offers functionality to infer peptide reactivity based on the edgeR package, though the improvement in speed is offset by slightly lower sensitivity compared to the Bayesian approach, specifically for weakly reactive peptides. AVAILABILITY AND IMPLEMENTATION: BEER is implemented in R and freely available from the Bioconductor repository at https://bioconductor.org/packages/release/bioc/html/beer.html.

Phage display of environmental protein toxins and virulence factors reveals the prevalence, persistence, and genetics of antibody responses.

Microbial exposures are crucial environmental factors that impact healthspan by sculpting the immune system and microbiota. Antibody profiling via Phage ImmunoPrecipitation Sequencing (PhIP-Seq) provides a high-throughput, cost-effective approach for detecting exposure and response to microbial protein products. We designed and constructed a library of 95,601 56-amino acid peptide tiles spanning 14,430 proteins with "toxin" or "virulence factor" keyword annotations. We used PhIP-Seq to profile the antibodies of ∼1,000 individuals against this "ToxScan" library. In addition to enumerating immunodominant antibody epitopes, we studied the age-dependent stability of the ToxScan profile and used a genome-wide association study to find that the MHC-II locus modulates bacterial epitope selection. We detected previously described anti-flagellin antibody responses in a Crohn's disease cohort and identified an association between anti-flagellin antibodies and juvenile dermatomyositis. PhIP-Seq with the ToxScan library is thus an effective tool for studying the environmental determinants of health and disease at cohort scale.

Polygenic transcriptome risk scores for COPD and lung function improve cross-ethnic portability of prediction in the NHLBI TOPMed program.

While polygenic risk scores (PRSs) enable early identification of genetic risk for chronic obstructive pulmonary disease (COPD), predictive performance is limited when the discovery and target populations are not well matched. Hypothesizing that the biological mechanisms of disease are shared across ancestry groups, we introduce a PrediXcan-derived polygenic transcriptome risk score (PTRS) to improve cross-ethnic portability of risk prediction. We constructed the PTRS using summary statistics from application of PrediXcan on large-scale GWASs of lung function (forced expiratory volume in 1 s [FEV1] and its ratio to forced vital capacity [FEV1/FVC]) in the UK Biobank. We examined prediction performance and cross-ethnic portability of PTRS through smoking-stratified analyses both on 29,381 multi-ethnic participants from TOPMed population/family-based cohorts and on 11,771 multi-ethnic participants from TOPMed COPD-enriched studies. Analyses were carried out for two dichotomous COPD traits (moderate-to-severe and severe COPD) and two quantitative lung function traits (FEV1 and FEV1/FVC). While the proposed PTRS showed weaker associations with disease than PRS for European ancestry, the PTRS showed stronger association with COPD than PRS for African Americans (e.g., odds ratio [OR] = 1.24 [95% confidence interval [CI]: 1.08-1.43] for PTRS versus 1.10 [0.96-1.26] for PRS among heavy smokers with ≥ 40 pack-years of smoking) for moderate-to-severe COPD. Cross-ethnic portability of the PTRS was significantly higher than the PRS (paired t test p < 2.2 × 10-16 with portability gains ranging from 5% to 28%) for both dichotomous COPD traits and across all smoking strata. Our study demonstrates the value of PTRS for improved cross-ethnic portability compared to PRS in predicting COPD risk.

Secondary analyses for genome-wide association studies using expression quantitative trait loci.

Genome-wide association studies (GWAS) have successfully identified thousands of single nucleotide polymorphisms (SNPs) associated with complex traits; however, the identified SNPs account for a fraction of trait heritability, and identifying the functional elements through which genetic variants exert their effects remains a challenge. Recent evidence suggests that SNPs associated with complex traits are more likely to be expression quantitative trait loci (eQTL). Thus, incorporating eQTL information can potentially improve power to detect causal variants missed by traditional GWAS approaches. Using genomic, transcriptomic, and platelet phenotype data from the Genetic Study of Atherosclerosis Risk family-based study, we investigated the potential to detect novel genomic risk loci by incorporating information from eQTL in the relevant target tissues (i.e., platelets and megakaryocytes) using established statistical principles in a novel way. Permutation analyses were performed to obtain family-wise error rates for eQTL associations, substantially lowering the genome-wide significance threshold for SNP-phenotype associations. In addition to confirming the well known association between PEAR1 and platelet aggregation, our eQTL-focused approach identified a novel locus (rs1354034) and gene (ARHGEF3) not previously identified in a GWAS of platelet aggregation phenotypes. A colocalization analysis showed strong evidence for a functional role of this eQTL.

HLA alleles and sustained peanut consumption promote IgG4 responses in subjects protected from peanut allergy.

We investigated the interplay between genetics and oral peanut protein exposure in the determination of the immunological response to peanut using the targeted intervention in the LEAP clinical trial. We identified an association between peanut-specific IgG4 and HLA-DQA1*01:02 that was only observed in the presence of sustained oral peanut protein exposure. The association between IgG4 and HLA-DQA1*01:02 was driven by IgG4 specific for the Ara h 2 component. Once peanut consumption ceased, the association between IgG4-specific Ara h 2 and HLA-DQA1*01:02 was attenuated. The association was validated by observing expanded IgG4-specific epitopes in people who carried HLA-DQA1*01:02. Notably, we confirmed the previously reported associations with HLA-DQA1*01:02 and peanut allergy risk in the absence of oral peanut protein exposure. Interaction between HLA and presence or absence of exposure to peanut in an allergen- and epitope-specific manner implicates a mechanism of antigen recognition that is fundamental to driving immune responses related to allergy risk or protection.

A pre-processing pipeline to quantify, visualize, and reduce technical variation in protein microarray studies.

Technical variation, or variation from non-biological sources, is present in most laboratory assays. Correcting for this variation enables analysts to extract a biological signal that informs questions of interest. However, each assay has different sources and levels of technical variation, and the choice of correction methods can impact downstream analyses. Compared to similar assays such as DNA microarrays, relatively few methods have been developed and evaluated for protein microarrays, a versatile tool for measuring levels of various proteins in serum samples. Here, we propose a pre-processing pipeline to correct for some common sources of technical variation in protein microarrays. The pipeline builds upon an existing normalization method by using controls to reduce technical variation. We evaluate our method using data from two protein microarray studies and by simulation. We demonstrate that pre-processing choices impact the fluorescent-intensity based ranks of proteins, which in turn, impact downstream analysis.

HIV Antibody Profiles in HIV Controllers and Persons With Treatment-Induced Viral Suppression.

Introduction: Low HIV viral load is associated with delayed disease progression and reduced HIV transmission. HIV controllers suppress viral load to low levels in the absence of antiretroviral treatment (ART). We used an antibody profiling system, VirScan, to compare antibody reactivity and specificity in HIV controllers, non-controllers with treatment-induced viral suppression, and viremic non-controllers. Methods: The VirScan library contains 3,384 phage-displayed peptides spanning the HIV proteome. Antibody reactivity to these peptides was measured in plasma from a Discovery Cohort that included 13 elite controllers, 27 viremic controllers, 12 viremic non-controllers, and 21 non-controllers who were virally suppressed on ART. Antibody reactivity to selected peptides was also assessed in an independent cohort of 29 elite controllers and 37 non-controllers who were virally suppressed on ART (Validation Cohort) and in a longitudinal cohort of non-controllers. Results: In the Discovery Cohort, 62 peptides were preferentially targeted in HIV controllers compared to non-controllers who were virally suppressed on ART. These specificities were not significantly different when comparing controllers versus viremic non-controllers. Aggregate reactivity to these peptides was also high in elite controllers from the independent Validation Cohort. The 62 peptides formed seven clusters of homologous epitopes in env, gag, integrase, and vpu. Reactivity to one of these clusters located in gag p17 was inversely correlated with viral load set point in an independent cohort of non-controllers. Conclusions: Antibody reactivity was low in non-controllers suppressed on ART, but remained high in viremic controllers despite viral suppression. Antibodies in controllers and viremic non-controllers were directed against epitopes in diverse HIV proteins; higher reactivity against p17 peptides was associated with lower viral load set point. Further studies are needed to determine if these antibodies play a role in regulation of HIV viral load.

Multiethnic genome-wide and HLA association study of total serum IgE level.

BACKGROUND: Total serum IgE (tIgE) is an important intermediate phenotype of allergic disease. Whole genome genetic association studies across ancestries may identify important determinants of IgE. OBJECTIVE: We aimed to increase understanding of genetic variants affecting tIgE production across the ancestry and allergic disease spectrum by leveraging data from the National Heart, Lung and Blood Institute Trans-Omics for Precision Medicine program; the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA); and the Atopic Dermatitis Research Network (N = 21,901). METHODS: We performed genome-wide association within strata of study, disease, and ancestry groups, and we combined results via a meta-regression approach that models heterogeneity attributable to ancestry. We also tested for association between HLA alleles called from whole genome sequence data and tIgE, assessing replication of associations in HLA alleles called from genotype array data. RESULTS: We identified 6 loci at genome-wide significance (P < 5 × 10-9), including 4 loci previously reported as genome-wide significant for tIgE, as well as new regions in chr11q13.5 and chr15q22.2, which were also identified in prior genome-wide association studies of atopic dermatitis and asthma. In the HLA allele association study, HLA-A∗02:01 was associated with decreased tIgE level (Pdiscovery = 2 × 10-4; Preplication = 5 × 10-4; Pdiscovery+replication = 4 × 10-7), and HLA-DQB1∗03:02 was strongly associated with decreased tIgE level in Hispanic/Latino ancestry populations (PHispanic/Latino discovery+replication = 8 × 10-8). CONCLUSION: We performed the largest genome-wide association study and HLA association study of tIgE focused on ancestrally diverse populations and found several known tIgE and allergic disease loci that are relevant in non-European ancestry populations.