Quality control and analytic best practices for testing genetic models of sex differences in large populations.

Phenotypic sex-based differences exist for many complex traits. In other cases, phenotypes may be similar, but underlying biology may vary. Thus, sex-aware genetic analyses are becoming increasingly important for understanding the mechanisms driving these differences. To this end, we provide a guide outlining the current best practices for testing various models of sex-dependent genetic effects in complex traits and disease conditions, noting that this is an evolving field. Insights from sex-aware analyses will not only teach us about the biology of complex traits but also aid in achieving the goals of precision medicine and health equity for all.

Bedside Back to Bench: Building Bridges between Basic and Clinical Genomic Research.

Genome sequencing has revolutionized the diagnosis of genetic diseases. Close collaborations between basic scientists and clinical genomicists are now needed to link genetic variants with disease causation. To facilitate such collaborations, we recommend prioritizing clinically relevant genes for functional studies, developing reference variant-phenotype databases, adopting phenotype description standards, and promoting data sharing.

The role of sex in the genomics of human complex traits.

Nearly all human complex traits and disease phenotypes exhibit some degree of sex differences, including differences in prevalence, age of onset, severity or disease progression. Until recently, the underlying genetic mechanisms of such sex differences have been largely unexplored. Advances in genomic technologies and analytical approaches are now enabling a deeper investigation into the effect of sex on human health traits. In this Review, we discuss recent insights into the genetic models and mechanisms that lead to sex differences in complex traits. This knowledge is critical for developing deeper insight into the fundamental biology of sex differences and disease processes, thus facilitating precision medicine.

Author Correction: The role of sex in the genomics of human complex traits.

In Box 4 of the originally published article, the text describing the Miami plot in part c of the figure contained some minor errors. Specifically, during pre-publication revision of the article, the authors updated the illustrative Miami plot (shown in figure part c) from that of reference 80 (Randall et al. PLoS Genet. (2013)) to a more recent study in reference 82 (Winkler et al. PLoS Genet. (2015)). The box text has now been updated to reflect that change. In paragraph 2, the trait has been updated from "hip circumference adjusted for body mass index" to "waist-to-hip ratio adjusted for body mass index (under 50 years old)" and it has been clarified that female GWAS data are shown on the top half of the plot with male data at the bottom. The original two citations of reference 80 in the Box 4 text have been updated to reference 82. Finally, a typographical artefact was corrected on the Y axis of the Miami plot, whereby the labels '14' and '16' in the top half of the plot we originally both shown as '12'. None of these corrections alter the overall illustrative point that genetic architectures for traits can differ between males and females, which was a conclusion of both reference 80 and reference 82.

A deep transcriptome meta-analysis reveals sex differences in multiple sclerosis.

BACKGROUND: Multiple sclerosis (MS), a chronic auto-immune, inflammatory, and degenerative disease of the central nervous system, affects both males and females; however, females suffer from a higher risk of developing MS (2-3:1 ratio relative to males). The precise sex-based factors influencing risk of MS are currently unknown. Here, we explore the role of sex in MS to identify molecular mechanisms underlying observed MS sex differences that may guide novel therapeutic approaches tailored for males or females. METHODS: We performed a rigorous and systematic review of genome-wide transcriptome studies of MS that included patient sex data in the Gene Expression Omnibus and ArrayExpress databases following PRISMA statement guidelines. For each selected study, we analyzed differential gene expression to explore the impact of the disease in females (IDF), in males (IDM) and our main goal: the sex differential impact of the disease (SDID). Then, for each scenario (IDF, IDM and SDID) we performed 2 meta-analyses in the main tissues involved in the disease (brain and blood). Finally, we performed a gene set analysis in brain tissue, in which a higher number of genes were dysregulated, to characterize sex differences in biological pathways. RESULTS: After screening 122 publications, the systematic review provided a selection of 9 studies (5 in blood and 4 in brain tissue) with a total of 474 samples (189 females with MS and 109 control females; 82 males with MS and 94 control males). Blood and brain tissue meta-analyses identified, respectively, 1 (KIR2DL3) and 13 (ARL17B, CECR7, CEP78, IFFO2, LOC401127, NUDT18, RNF10, SLC17A5, STMP1, TRAF3IP2-AS1, UBXN2B, ZNF117, ZNF488) MS-associated genes that differed between males and females (SDID comparison). Functional analyses in the brain revealed different altered immune patterns in females and males (IDF and IDM comparisons). The pro-inflammatory environment and innate immune responses related to myeloid lineage appear to be more affected in females, while adaptive responses associated with the lymphocyte lineage in males. Additionally, females with MS displayed alterations in mitochondrial respiratory chain complexes, purine, and glutamate metabolism, while MS males displayed alterations in stress response to metal ion, amine, and amino acid transport. CONCLUSION: We found transcriptomic and functional differences between MS males and MS females (especially in the immune system), which may support the development of new sex-based research of this disease. Our study highlights the importance of understanding the role of biological sex in MS to guide a more personalized medicine.

Systematic evaluation of transcriptomics-based deconvolution methods and references using thousands of clinical samples.

Estimating cell type composition of blood and tissue samples is a biological challenge relevant in both laboratory studies and clinical care. In recent years, a number of computational tools have been developed to estimate cell type abundance using gene expression data. Although these tools use a variety of approaches, they all leverage expression profiles from purified cell types to evaluate the cell type composition within samples. In this study, we compare 12 cell type quantification tools and evaluate their performance while using each of 10 separate reference profiles. Specifically, we have run each tool on over 4000 samples with known cell type proportions, spanning both immune and stromal cell types. A total of 12 of these represent in vitro synthetic mixtures and 300 represent in silico synthetic mixtures prepared using single-cell data. A final 3728 clinical samples have been collected from the Framingham cohort, for which cell populations have been quantified using electrical impedance cell counting. When tools are applied to the Framingham dataset, the tool Estimating the Proportions of Immune and Cancer cells (EPIC) produces the highest correlation, whereas Gene Expression Deconvolution Interactive Tool (GEDIT) produces the lowest error. The best tool for other datasets is varied, but CIBERSORT and GEDIT most consistently produce accurate results. We find that optimal reference depends on the tool used, and report suggested references to be used with each tool. Most tools return results within minutes, but on large datasets runtimes for CIBERSORT can exceed hours or even days. We conclude that deconvolution methods are capable of returning high-quality results, but that proper reference selection is critical.

The Evolutionary History of Common Genetic Variants Influencing Human Cortical Surface Area.

Structural brain changes along the lineage leading to modern Homo sapiens contributed to our distinctive cognitive and social abilities. However, the evolutionarily relevant molecular variants impacting key aspects of neuroanatomy are largely unknown. Here, we integrate evolutionary annotations of the genome at diverse timescales with common variant associations from large-scale neuroimaging genetic screens. We find that alleles with evidence of recent positive polygenic selection over the past 2000-3000 years are associated with increased surface area (SA) of the entire cortex, as well as specific regions, including those involved in spoken language and visual processing. Therefore, polygenic selective pressures impact the structure of specific cortical areas even over relatively recent timescales. Moreover, common sequence variation within human gained enhancers active in the prenatal cortex is associated with postnatal global SA. We show that such variation modulates the function of a regulatory element of the developmentally relevant transcription factor HEY2 in human neural progenitor cells and is associated with structural changes in the inferior frontal cortex. These results indicate that non-coding genomic regions active during prenatal cortical development are involved in the evolution of human brain structure and identify novel regulatory elements and genes impacting modern human brain structure.

Sex differences in gene expression in response to ischemia in the human left ventricular myocardium.

Sex differences exist in the prevalence, presentation and outcomes of ischemic heart disease (IHD). Females have higher risk of heart failure post-myocardial infarction relative to males and are two to three times more likely to die after coronary artery bypass grafting surgery. We examined sex differences in human myocardial gene expression in response to ischemia. Left ventricular biopsies from 68 male/46 female patients undergoing aortic valve replacement surgery were obtained at baseline and after a median 74 min of cold cardioplegic arrest/ischemia. Transcriptomes were quantified by RNA-sequencing. Cell-type enrichment analysis was used to estimate the identity and relative proportions of different cell types in each sample. A sex-specific response to ischemia was observed for 271 genes. Notably, the expression FAM5C, PLA2G4E and CYP1A1 showed an increased expression in females compared to males due to ischemia and DIO3, MT1G and CMA1 showed a decreased expression in females compared to males due to ischemia. Functional annotation analysis revealed sex-specific modulation of the oxytocin signaling pathway and common pathway of fibrin clot formation. Expression quantitative trait locus (eQTL) analysis identified variant-by-sex interaction eQTLs, indicative of sex differences in the genotypic effects on gene expression. Cell-type enrichment analysis showed sex-bias in proportion of specific cell types. Common lymphoid progenitor cells and M2 macrophages were found to increase in female samples from pre- to post-ischemia, but no change was observed in male samples. These differences in response to myocardial ischemia provide insight into the sexual dimorphism of IHD and may aid in the development of sex-specific therapies that reduce myocardial injury.

Genetic analysis of isoform usage in the human anti-viral response reveals influenza-specific regulation of ERAP2 transcripts under balancing selection.

While genetic variants are known to be associated with overall gene abundance in stimulated immune cells, less is known about their effects on alternative isoform usage. By analyzing RNA-seq profiles of monocyte-derived dendritic cells from 243 individuals, we uncovered thousands of unannotated isoforms synthesized in response to influenza infection and type 1 interferon stimulation. We identified more than a thousand quantitative trait loci (QTLs) associated with alternate isoform usage (isoQTLs), many of which are independent of expression QTLs (eQTLs) for the same gene. Compared with eQTLs, isoQTLs are enriched for splice sites and untranslated regions, but depleted of sequences upstream of annotated transcription start sites. Both eQTLs and isoQTLs explain a significant proportion of the disease heritability attributed to common genetic variants. At the ERAP2 locus, we shed light on the function of the gene and how two frequent, highly differentiated haplotypes with intermediate frequencies could be maintained by balancing selection. At baseline and following type 1 interferon stimulation, the major haplotype is associated with low ERAP2 expression caused by nonsense-mediated decay, while the minor haplotype, known to increase Crohn's disease risk, is associated with high ERAP2 expression. In response to influenza infection, we found two uncharacterized isoforms expressed from the major haplotype, likely the result of multiple perfectly linked variants affecting the transcription and splicing at the locus. Thus, genetic variants at a single locus could modulate independent gene regulatory processes in innate immune responses and, in the case of ERAP2, may confer a historical fitness advantage in response to virus.

Large-Scale trans-eQTLs Affect Hundreds of Transcripts and Mediate Patterns of Transcriptional Co-regulation.

Efforts to decipher the causal relationships between differences in gene regulation and corresponding differences in phenotype have been stymied by several basic technical challenges. Although detecting local, cis-eQTLs is now routine, trans-eQTLs, which are distant from the genes of origin, are far more difficult to find because millions of SNPs must currently be compared to thousands of transcripts. Here, we demonstrate an alternative approach: we looked for SNPs associated with the expression of many genes simultaneously and found that hundreds of trans-eQTLs each affect hundreds of transcripts in lymphoblastoid cell lines across three African populations. These trans-eQTLs target the same genes across the three populations and show the same direction of effect. We discovered that target transcripts of a high-confidence set of trans-eQTLs encode proteins that interact more frequently than expected by chance, are bound by the same transcription factors, and are enriched for pathway annotations indicative of roles in basic cell homeostasis. We thus demonstrate that our approach can uncover trans-acting transcriptional control circuits that affect co-regulated groups of genes: a key to understanding how cellular pathways and processes are orchestrated.

Pharmacogenomic genotypes define genetic ancestry in patients and enable population-specific genomic implementation.

The importance of genetic ancestry characterization is increasing in genomic implementation efforts, and clinical pharmacogenomic guidelines are being published that include population-specific recommendations. Our aim was to test the ability of focused clinical pharmacogenomic SNP panels to estimate individual genetic ancestry (IGA) and implement population-specific pharmacogenomic clinical decision-support (CDS) tools. Principle components and STRUCTURE were utilized to assess differences in genetic composition and estimate IGA among 1572 individuals from 1000 Genomes, two independent cohorts of Caucasians and African Americans (AAs), plus a real-world validation population of patients undergoing pharmacogenomic genotyping. We found that clinical pharmacogenomic SNP panels accurately estimate IGA compared to genome-wide genotyping and identify AAs with ≥70 African ancestry (sensitivity >82%, specificity >80%, PPV >95%, NPV >47%). We also validated a new AA-specific warfarin dosing algorithm for patients with ≥70% African ancestry and implemented it at our institution as a novel CDS tool. Consideration of IGA to develop an institutional CDS tool was accomplished to enable population-specific pharmacogenomic guidance at the point-of-care. These capabilities were immediately applied for guidance of warfarin dosing in AAs versus Caucasians, but also provide a real-world model that can be extended to other populations and drugs as actionable genomic evidence accumulates.

Imputing Gene Expression in Uncollected Tissues Within and Beyond GTEx.

Gene expression and its regulation can vary substantially across tissue types. In order to generate knowledge about gene expression in human tissues, the Genotype-Tissue Expression (GTEx) program has collected transcriptome data in a wide variety of tissue types from post-mortem donors. However, many tissue types are difficult to access and are not collected in every GTEx individual. Furthermore, in non-GTEx studies, the accessibility of certain tissue types greatly limits the feasibility and scale of studies of multi-tissue expression. In this work, we developed multi-tissue imputation methods to impute gene expression in uncollected or inaccessible tissues. Via simulation studies, we showed that the proposed methods outperform existing imputation methods in multi-tissue expression imputation and that incorporating imputed expression data can improve power to detect phenotype-expression correlations. By analyzing data from nine selected tissue types in the GTEx pilot project, we demonstrated that harnessing expression quantitative trait loci (eQTLs) and tissue-tissue expression-level correlations can aid imputation of transcriptome data from uncollected GTEx tissues. More importantly, we showed that by using GTEx data as a reference, one can impute expression levels in inaccessible tissues in non-GTEx expression studies.

ImmVar project: Insights and design considerations for future studies of "healthy" immune variation.

The Immune Variation (ImmVar) project is one of a series of recent efforts to map the extent of variation in immune function in healthy human subjects. The focus of our initial studies involved a careful mapping of the genetic architecture of the adaptive and innate immunologic transcriptomes. Our studies highlight the shared nature of this immunogenetic architecture across human populations, the important role of context in uncovering effects of genetic variation, and the fact that, over all tested genes, common genetic variation account for a minority of the variance in the immune transcriptome in healthy subjects. Yet, it is an element of the variance that can be measured very precisely and will play an important role in the design of future studies. We therefore discuss how insights from ImmVar and similar studies inform experimental strategies and frame the design of future studies of immune function in health and disease.

Sex differences in the genetic architecture of obsessive-compulsive disorder.

Obsessive-compulsive disorder (OCD) is a highly heritable complex phenotype that demonstrates sex differences in age of onset and clinical presentation, suggesting a possible sex difference in underlying genetic architecture. We present the first genome-wide characterization of the sex-specific genetic architecture of OCD, utilizing the largest set of OCD cases and controls available from the Psychiatric Genomics Consortium. We assessed evidence for several mechanisms that may contribute to sex differences including a sex-dependent liability threshold, the presence of individual sex-specific risk variants on the autosomes and the X chromosome, and sex-specific pleiotropic effects. Furthermore, we tested the hypothesis that genetic heterogeneity between the sexes may obscure associations in a sex-combined genome-wide association study. We observed a strong genetic correlation between male and female OCD and no evidence for a sex-dependent liability threshold model, suggesting that sex-combined analysis does not suffer from widespread loss of power because of genetic heterogeneity between the sexes. While we did not detect any significant sex-specific genome-wide single nucleotide polymorphisms (SNP) associations, we did identify two significant gene-based associations in females: GRID2 and GRP135, which showed no association in males. We observed that the SNPs with sexually differentiated effects showed an enrichment of regulatory variants influencing expression of genes in brain and immune tissues. These findings suggest that future studies with larger sample sizes hold great promise for the identification of sex-specific genetic risk factors for OCD.

Disentangling the Effects of Colocalizing Genomic Annotations to Functionally Prioritize Non-coding Variants within Complex-Trait Loci.

Identifying genomic annotations that differentiate causal from trait-associated variants is essential to fine mapping disease loci. Although many studies have identified non-coding functional annotations that overlap disease-associated variants, these annotations often colocalize, complicating the ability to use these annotations for fine mapping causal variation. We developed a statistical approach (Genomic Annotation Shifter [GoShifter]) to assess whether enriched annotations are able to prioritize causal variation. GoShifter defines the null distribution of an annotation overlapping an allele by locally shifting annotations; this approach is less sensitive to biases arising from local genomic structure than commonly used enrichment methods that depend on SNP matching. Local shifting also allows GoShifter to identify independent causal effects from colocalizing annotations. Using GoShifter, we confirmed that variants in expression quantitative trail loci drive gene-expression changes though DNase-I hypersensitive sites (DHSs) near transcription start sites and independently through 3' UTR regulation. We also showed that (1) 15%-36% of trait-associated loci map to DHSs independently of other annotations; (2) loci associated with breast cancer and rheumatoid arthritis harbor potentially causal variants near the summits of histone marks rather than full peak bodies; (3) variants associated with height are highly enriched in embryonic stem cell DHSs; and (4) we can effectively prioritize causal variation at specific loci.

Accurate and fast multiple-testing correction in eQTL studies.

In studies of expression quantitative trait loci (eQTLs), it is of increasing interest to identify eGenes, the genes whose expression levels are associated with variation at a particular genetic variant. Detecting eGenes is important for follow-up analyses and prioritization because genes are the main entities in biological processes. To detect eGenes, one typically focuses on the genetic variant with the minimum p value among all variants in cis with a gene and corrects for multiple testing to obtain a gene-level p value. For performing multiple-testing correction, a permutation test is widely used. Because of growing sample sizes of eQTL studies, however, the permutation test has become a computational bottleneck in eQTL studies. In this paper, we propose an efficient approach for correcting for multiple testing and assess eGene p values by utilizing a multivariate normal distribution. Our approach properly takes into account the linkage-disequilibrium structure among variants, and its time complexity is independent of sample size. By applying our small-sample correction techniques, our method achieves high accuracy in both small and large studies. We have shown that our method consistently produces extremely accurate p values (accuracy > 98%) for three human eQTL datasets with different sample sizes and SNP densities: the Genotype-Tissue Expression pilot dataset, the multi-region brain dataset, and the HapMap 3 dataset.

Assocplots: a Python package for static and interactive visualization of multiple-group GWAS results.

Background: Over the last decade, genome-wide association studies (GWAS) have generated vast amounts of analysis results, requiring development of novel tools for data visualization. Quantile­quantile (QQ) plots and Manhattan plots are classical tools which have been utilized to visually summarize GWAS results and identify genetic variants significantly associated with traits of interest. However, static visualizations are limiting in the information that can be shown. Here, we present Assocplots, a Python package for viewing and exploring GWAS results not only using classic static Manhattan and QQ plots, but also through a dynamic extension which allows to interactively visualize the relationships between GWAS results from multiple cohorts or studies. Availability and Implementation: The Assocplots package is open source and distributed under the MIT license via GitHub (https://github.com/khramts/assocplots) along with examples, documentation and installation instructions. Contact: ekhramts@medicine.bsd.uchicago.edu or bstranger@medicine.bsd.uchicago.edu

Interindividual variation in human T regulatory cells.

FOXP3(+) regulatory T (Treg) cells enforce immune self-tolerance and homeostasis, and variation in some aspects of Treg function may contribute to human autoimmune diseases. Here, we analyzed population-level Treg variability by performing genome-wide expression profiling of CD4(+) Treg and conventional CD4(+) T (Tconv) cells from 168 donors, healthy or with established type-1 diabetes (T1D) or type-2 diabetes (T2D), in relation to genetic and immunologic screening. There was a range of variability in Treg signature transcripts, some almost invariant, others more variable, with more extensive variability for genes that control effector function (ENTPD1, FCRL1) than for lineage-specification factors like FOXP3 or IKZF2. Network analysis of Treg signature genes identified coregulated clusters that respond similarly to genetic and environmental variation in Treg and Tconv cells, denoting qualitative differences in otherwise shared regulatory circuits whereas other clusters are coregulated in Treg, but not Tconv, cells, suggesting Treg-specific regulation of genes like CTLA4 or DUSP4. Dense genotyping identified 110 local genetic variants (cis-expression quantitative trait loci), some of which are specifically active in Treg, but not Tconv, cells. The Treg signature became sharper with age and with increasing body-mass index, suggesting a tuning of Treg function with repertoire selection and/or chronic inflammation. Some Treg signature transcripts correlated with FOXP3 mRNA and/or protein, suggesting transcriptional or posttranslational regulatory relationships. Although no single transcript showed significant association to diabetes, overall expression of the Treg signature was subtly perturbed in T1D, but not T2D, patients.

Regulation of gene expression in autoimmune disease loci and the genetic basis of proliferation in CD4+ effector memory T cells.

Genome-wide association studies (GWAS) and subsequent dense-genotyping of associated loci identified over a hundred single-nucleotide polymorphism (SNP) variants associated with the risk of rheumatoid arthritis (RA), type 1 diabetes (T1D), and celiac disease (CeD). Immunological and genetic studies suggest a role for CD4-positive effector memory T (CD+ TEM) cells in the pathogenesis of these diseases. To elucidate mechanisms of autoimmune disease alleles, we investigated molecular phenotypes in CD4+ effector memory T cells potentially affected by these variants. In a cohort of genotyped healthy individuals, we isolated high purity CD4+ TEM cells from peripheral blood, then assayed relative abundance, proliferation upon T cell receptor (TCR) stimulation, and the transcription of 215 genes within disease loci before and after stimulation. We identified 46 genes regulated by cis-acting expression quantitative trait loci (eQTL), the majority of which we detected in stimulated cells. Eleven of the 46 genes with eQTLs were previously undetected in peripheral blood mononuclear cells. Of 96 risk alleles of RA, T1D, and/or CeD in densely genotyped loci, eleven overlapped cis-eQTLs, of which five alleles completely explained the respective signals. A non-coding variant, rs389862A, increased proliferative response (p=4.75 × 10-8). In addition, baseline expression of seventeen genes in resting cells reliably predicted proliferative response after TCR stimulation. Strikingly, however, there was no evidence that risk alleles modulated CD4+ TEM abundance or proliferation. Our study underscores the power of examining molecular phenotypes in relevant cells and conditions for understanding pathogenic mechanisms of disease variants.

CD33: increased inclusion of exon 2 implicates the Ig V-set domain in Alzheimer's disease susceptibility.

We previously demonstrated that the Alzheimer's disease (AD) associated risk allele, rs3865444(C), results in a higher surface density of CD33 on monocytes. Here, we find alternative splicing of exon 2 to be the primary mechanism of the genetically driven differential expression of CD33 protein. We report that the risk allele, rs3865444(C), is associated with greater cell surface expression of CD33 in both subjects of European and African-American ancestry and that there is a single haplotype influencing CD33 surface expression. A meta-analysis of the two populations narrowed the number of significant SNPs in high linkage disequilibrium (LD) (r(2) > 0.8) with rs3865444 to just five putative causal variants associated with increased protein expression. Using gene expression data from flow-sorted CD14(+)CD16(-) monocytes from 398 healthy subjects of three populations, we show that the rs3865444(C) risk allele is strongly associated with greater expression of CD33 exon 2 (pMETA = 2.36 × 10(-60)). Western blotting confirms increased protein expression of the full-length CD33 isoform containing exon 2 relative to the rs3865444(C) allele (P < 0.0001). Of the variants in strong LD with rs3865444, rs12459419, which is located in a putative SRSF2 splice site of exon 2, is the most likely candidate to mediate the altered alternative splicing of CD33's Immunoglobulin V-set domain 2 and ultimately influence AD susceptibility.