Complex Linkage Disequilibrium Research Articles

Genotype imputation methods for whole and complex genomic regions utilizing deep learning technology.

The imputation of unmeasured genotypes is essential in human genetic research, particularly in enhancing the power of genome-wide association studies and conducting subsequent fine-mapping. Recently, several deep learning-based genotype imputation methods for genome-wide variants with the capability of learning complex linkage disequilibrium patterns have been developed. Additionally, deep learning-based imputation has been applied to a distinct genomic region known as the major histocompatibility complex, referred to as HLA imputation. Despite their various advantages, the current deep learning-based genotype imputation methods do have certain limitations and have not yet become standard. These limitations include the modest accuracy improvement over statistical and conventional machine learning-based methods. However, their benefits include other aspects, such as their "reference-free" nature, which ensures complete privacy protection, and their higher computational efficiency. Furthermore, the continuing evolution of deep learning technologies is expected to contribute to further improvements in prediction accuracy and usability in the future.

HLA allele-calling using multi-ancestry whole-exome sequencing from the UK Biobank identifies 129 novel associations in 11 autoimmune diseases

The human leukocyte antigen (HLA) region on chromosome 6 is strongly associated with many immune-mediated and infection-related diseases. Due to its highly polymorphic nature and complex linkage disequilibrium patterns, traditional genetic association studies of single nucleotide polymorphisms do not perform well in this region. Instead, the field has adopted the assessment of the association of HLA alleles (i.e., entire HLA gene haplotypes) with disease. Often based on genotyping arrays, these association studies impute HLA alleles, decreasing accuracy and thus statistical power for rare alleles and in non-European ancestries. Here, we use whole-exome sequencing (WES) from 454,824 UK Biobank (UKB) participants to directly call HLA alleles using the HLA-HD algorithm. We show this method is more accurate than imputing HLA alleles and harness the improved statistical power to identify 360 associations for 11 auto-immune phenotypes (at least 129 likely novel), leading to better insights into the specific coding polymorphisms that underlie these diseases. We show that HLA alleles with synonymous variants, often overlooked in HLA studies, can significantly influence these phenotypes. Lastly, we show that HLA sequencing may improve polygenic risk scores accuracy across ancestries. These findings allow better characterization of the role of the HLA region in human disease.

Coexistence of Multiple Functional Variants and Genes Underlies Genetic Risk Locus 11p11.2 of Alzheimer’s Disease

BackgroundGenome-wide association studies have identified dozens of genetic risk loci for Alzheimer’s disease (AD), yet the underlying causal variants and biological mechanisms remain elusive, especially for loci with complex linkage disequilibrium and regulation. MethodsTo fully untangle the causal signal at a single locus, we performed a functional genomic study of 11p11.2 (the CELF1/SPI1 locus). Genome-wide association study signals at 11p11.2 were integrated with datasets of histone modification, open chromatin, and transcription factor binding to distill potentially functional variants (fVars). Their allelic regulatory activities were confirmed by allele imbalance, reporter assays, and base editing. Expressional quantitative trait loci and chromatin interaction data were incorporated to assign target genes to fVars. The relevance of these genes to AD was assessed by convergent functional genomics using bulk brain and single-cell transcriptomic, epigenomic, and proteomic datasets of patients with AD and control individuals, followed by cellular assays. ResultsWe found that 24 potential fVars, rather than a single variant, were responsible for the risk of 11p11.2. These fVars modulated transcription factor binding and regulated multiple genes by long-range chromatin interactions. Besides SPI1, convergent evidence indicated that 6 target genes (MTCH2, ACP2, NDUFS3, PSMC3, C1QTNF4, and MADD) of fVars were likely to be involved in AD development. Disruption of each gene led to cellular amyloid-β and phosphorylated tau changes, supporting the existence of multiple likely causal genes at 11p11.2. ConclusionsMultiple variants and genes at 11p11.2 may contribute to AD risk. This finding provides new insights into the mechanistic and therapeutic challenges of AD.

Investigation of the causal relationship between ALS and autoimmune disorders: a Mendelian randomization study

BackgroundEpidemiological studies have reported an association between amyotrophic lateral sclerosis (ALS) and different autoimmune disorders. This study aims to explore the causal relationship between autoimmune disorders and ALS using Mendelian randomization (MR).MethodsTo test the genetically predicted effects of liability towards immune-related outcomes on ALS risk, we used summary statistics from the largest European genome-wide association studies (GWAS) for these disorders in a two-sample MR setting. To do this, we extracted single nucleotide polymorphisms (SNPs) from the GWAS, which strongly associated with the 12 traits, and queried their effects in a large European ALS GWAS (27,265 cases and 110,881 controls). To avoid bias in our MR instruments related to the complex linkage disequilibrium structure of the human leukocyte antigen (HLA) region, we excluded SNPs within this region from the analyses. We computed inverse-variance weighted (IVW) MR estimates and undertook sensitivity analyses using MR methods robust to horizontal pleiotropy. We also performed a reverse MR analysis testing the causal effects of ALS on the above autoimmune traits.ResultsAfter applying Bonferroni correction for multiple testing, our MR analyses showed that the liability to autoimmune disorders does not affect ALS risk. Our reverse MR analysis also did not support the effects of liability to ALS on other autoimmune disorders. The results of the main IVW MR analyses were generally supported by our sensitivity MR analyses. The variance in the exposures explained by the sets of SNPs used as MR instruments ranged from 8.1 × 10−4 to 0.31. Our MR study was well-powered to detect effects as small as an odds ratio (OR) of 1.045 for ALS in the main MR and as small as an OR of 1.32 in the reverse MR.ConclusionOur MR study does not support a relationship between liability to autoimmune disorders and ALS risk in the European population. The associations observed in epidemiological studies could be partly attributed to shared biology or environmental confounders.

Rapid, Reference-Free human genotype imputation with denoising autoencoders.

Genotype imputation is a foundational tool for population genetics. Standard statistical imputation approaches rely on the co-location of large whole-genome sequencing-based reference panels, powerful computing environments, and potentially sensitive genetic study data. This results in computational resource and privacy-risk barriers to access to cutting-edge imputation techniques. Moreover, the accuracy of current statistical approaches is known to degrade in regions of low and complex linkage disequilibrium. Artificial neural network-based imputation approaches may overcome these limitations by encoding complex genotype relationships in easily portable inference models. Here, we demonstrate an autoencoder-based approach for genotype imputation, using a large, commonly used reference panel, and spanning the entirety of human chromosome 22. Our autoencoder-based genotype imputation strategy achieved superior imputation accuracy across the allele-frequency spectrum and across genomes of diverse ancestry, while delivering at least fourfold faster inference run time relative to standard imputation tools.

Coding and Noncoding Variation in LRRK2 and Parkinson's Disease Risk.

BackgroundThe leucine‐rich repeat kinase 2 (LRRK2) gene harbors both rare highly damaging missense variants (eg, p.G2019S) and common noncoding variants (eg, rs76904798) with lower effect sizes that are associated with Parkinson's disease (PD) risk.ObjectivesThis study aimed to investigate in a large meta‐analysis whether the LRRK2 Genome‐Wide Association Study (GWAS) signal represented by rs76904798 is independently associated with PD risk from LRRK2 coding variation and whether complex linkage disequilibrium structures with p.G2019S and the 5′ noncoding haplotype account for the association of LRRK2 coding variants.MethodsWe performed a meta‐analysis using imputed genotypes from 17,838 patients, 13,404 proxy patients, and 173,639 healthy controls of European ancestry. We excluded carriers of p.G2019S and/or rs76904798 to clarify the role of LRRK2 coding variation in mediating disease risk and excluded carriers of relatively rare LRRK2 coding variants to assess the independence of rs76904798. We also investigated the co‐inheritance of LRRK2 coding variants with p.G2019S, rs76904798, and p.N2081D.Results LRRK2 rs76904798 remained significantly associated with PD after excluding the carriers of relatively rare LRRK2 coding variants. LRRK2 p.R1514Q and p.N2081D were frequently co‐inherited with rs76904798, and the allele distribution of p.S1647T significantly changed among patients after removing rs76904798 carriers.ConclusionsThese data suggest that the LRRK2 coding variants previously related to PD (p.N551K, p.R1398H, p.M1646T, and p.N2081D) do not drive the 5′ noncoding GWAS signal. These data, however, do not preclude the independent association of the haplotype p.N551K‐p.R1398H and p.M1646T with altered disease risk. © 2021 The Authors. Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson Movement Disorder Society. This article has been contributed to by US Government employees and their work is in the public domain in the USA.

Fine mapping of the major histocompatibility complex (MHC) in myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS) suggests involvement of both HLA class I and class II loci

The etiology of myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS) is unknown, but involvement of the immune system is one of the proposed underlying mechanisms. Human leukocyte antigen (HLA) associations are hallmarks of immune-mediated and autoimmune diseases. We have previously performed high resolution HLA genotyping and detected associations between ME/CFS and certain HLA class I and class II alleles. However, the HLA complex harbors numerous genes of immunological importance, and there is extensive and complex linkage disequilibrium across the region. In the current study, we aimed to fine map the association signals in the HLA complex by genotyping five additional classical HLA loci and 5,342 SNPs in 427 Norwegian ME/CFS patients, diagnosed according to the Canadian Consensus Criteria, and 480 healthy Norwegian controls. SNP association analysis revealed two distinct and independent association signals (p ≤ 0.001) tagged by rs4711249 in the HLA class I region and rs9275582 in the HLA class II region. Furthermore, the primary association signal in the HLA class II region was located within the HLA-DQ gene region, most likely due to HLA-DQB1, particularly the amino acid position 57 (aspartic acid/alanine) in the peptide binding groove, or an intergenic SNP upstream of HLA-DQB1. In the HLA class I region, the putative causal locus might map outside the classical HLA genes as the association signal spans several genes (DDR1, GTF2H4, VARS2, SFTA2 and DPCR1) with expression levels influenced by the ME/CFS associated SNP genotype. Taken together, our results implicate the involvement of the MHC, and in particular the HLA-DQB1 gene, in ME/CFS. These findings should be replicated in larger cohorts, particularly to verify the putative involvement of HLA-DQB1, a gene important for antigen-presentation to T cells and known to harbor alleles providing the largest risk for well–established autoimmune diseases.

245-OR: Comprehensive Genome-Wide Association Study (GWAS) Meta-analysis Using TOPMed Imputation in Latinos Identifies Rare Variation Associated with Type 2 Diabetes (T2D)

The high prevalence of T2D among Latino individuals emphasizes the importance of studying both environmental and genetic factors that influence disease in this population. However, Latino is a recently admixed population with complex linkage disequilibrium patterns that impose methodological challenges, especially when using genotype imputation to test association with rare variation. The recent release of the multi-ethnic NHLBI Trans-Omics for Precision Medicine (TOPMed) reference panel, comprising more than 100,000 samples (~10% are Latinos), represents a unique opportunity to explore rare genetic variation in this population. We performed the largest T2D GWAS meta-analysis in the Latino population imputed to the TOPMed panel. Our analysis includes 18,294 individuals (7,556 cases and 10,738 controls) with genotyping array and whole-exome sequencing data. We were able to impute 38M high-quality variants (rsq>0.8) with TOPMed (5M with minor allele frequency [MAF] 0.001-0.0005) compared to 16M with 1000G Phase 3 panel (1.6M with [MAF] 0.001-0.0005). This allowed us to replicate previously reported T2D-associated common variants, and to identify a novel genome-wide significant variant (MAF=0.016, OR=2.06, P=1.7×10-9) near ORC5, a gene implicated in familial hyperinsulinemic hypoglycemia 2. Two additional suggestive novel signals merit further research: a variant in ROR2 (MAF=0.066, OR=1.3, P=8.8×10-8) and a Latino specific variant in HDAC2 only identified with the use of TOPMed (MAF=0.013, OR=1.82, P=3.2×10-7) and associated with insulin secretion in individuals without diabetes in this study. Our results demonstrate the utility of the TOPMed imputation for the identification of novel and population-specific rare variation associated with T2D in understudied populations, which could lead to the identification of novel therapeutic targets and the improvement of polygenic risk scores. Disclosure A. Huerta: None. Mgb biobank: n/a. J. B. Cole: None. P. Dornbos: None. D. A. Dicorpo: None. A. Leong: None. J. C. Florez: Consultant; Self; Goldfinch Bio, Inc., Other Relationship; Self; Novo Nordisk. J. M. Mercader: None. Sigma t2d consortium: n/a. Mx biobank: n/a. Funding American Diabetes Association (1-19-ICTS-068 to J.M.M.)

Genome-wide haplotype association study in imaging genetics using whole-brain sulcal openings of 16,304 UK Biobank subjects

Neuroimaging-genetics cohorts gather two types of data: brain imaging and genetic data. They allow the discovery of associations between genetic variants and brain imaging features. They are invaluable resources to study the influence of genetics and environment in the brain features variance observed in normal and pathological populations. This study presents a genome-wide haplotype analysis for 123 brain sulcus opening value (a measure of sulcal width) across the whole brain that include 16,304 subjects from UK Biobank. Using genetic maps, we defined 119,548 blocks of low recombination rate distributed along the 22 autosomal chromosomes and analyzed 1,051,316 haplotypes. To test associations between haplotypes and complex traits, we designed three statistical approaches. Two of them use a model that includes all the haplotypes for a single block, while the last approach considers each haplotype independently. All the statistics produced were assessed as rigorously as possible. Thanks to the rich imaging dataset at hand, we used resampling techniques to assess False Positive Rate for each statistical approach in a genome-wide and brain-wide context. The results on real data show that genome-wide haplotype analyses are more sensitive than single-SNP approach and account for local complex Linkage Disequilibrium (LD) structure, which makes genome-wide haplotype analysis an interesting and statistically sound alternative to the single-SNP counterpart.

Complex Linkage Disequilibrium Effects in HLA-DPB1 Expression and Molecular Mismatch Analyses of Transplantation Outcomes.

HLA molecular mismatch (MM) is a risk factor for de novo donor-specific antibody (dnDSA) development in solid organ transplantation. HLA expression differences have also been associated with adverse outcomes in hematopoietic cell transplantation. We sought to study both MM and expression in assessing dnDSA risk. One hundred three HLA-DP-mismatched solid organ transplantation pairs were retrospectively analyzed. MM was computed using amino acids (aa), eplets, and, supplementarily, Grantham/Epstein scores. DPB1 alleles were classified as rs9277534-A (low-expression) or rs9277534-G (high-expression) linked. To determine the associations between risk factors and dnDSA, logistic regression, linkage disequilibrium (LD), and population-based analyses were performed. A high-risk AA:GX (recipient:donor) expression combination (X = A or G) demonstrated strong association with HLA-DP dnDSA (P = 0.001). MM was also associated with HLA-DP dnDSA when evaluated by itself (eplet P = 0.007, aa P = 0.003, Grantham P = 0.005, Epstein P = 0.004). When attempting to determine the relative individual effects of the risk factors in multivariable analysis, only AA:GX expression status retained a strong association (relative risk = 18.6, P = 0.007 with eplet; relative risk = 15.8, P = 0.02 with aa), while MM was no longer significant (eplet P = 0.56, aa P = 0.51). Importantly, these risk factors are correlated, due to LD between the expression-tagging single-nucleotide polymorphism and polymorphisms along HLA-DPB1. The MM and expression risk factors each appear to be strong predictors of HLA-DP dnDSA and to possess clinical utility; however, these two risk factors are closely correlated. These metrics may represent distinct ways of characterizing a common overlapping dnDSA risk profile, but they are not independent. Further, we demonstrate the importance and detailed implications of LD effects in dnDSA risk assessment and possibly transplantation overall.

Genome‐wide profiling of the noncoding regulatory mechanisms in Alzheimer’s disease

AbstractBackgroundGenome‐wide association studies (GWAS) have identified genetic variants associated with Alzheimer’s Disease (AD), but the majority are noncoding and may affect transcriptional regulatory elements such as enhancers. While prior post‐GWAS analyses focused on analyzing only top GWAS signals, this study focuses on a systematic genome‐wide analysis of all significant GWAS signals aimed at identifying a more complete set of causal variants given a complex linkage disequilibrium (LD) genome structures and the regulatory mechanisms, and tissues/cell types.MethodTo perform genome‐wide analysis of GWAS summary statistics (NIAGADS NG00036) from the IGAP study [Lambert et al., 2013], we adopted INFERNO (http://inferno.lisanwanglab.org), a recent method for post‐GWAS analysis of non‐coding variants. INFERNO integrates GWAS data with hundreds of publicly available functional genomics datasets (e.g., ENCODE, FANTOM5, GTEx, HOMER) across 485 tissues and cell types to facilitate identification of the regulatory mechanisms and tissue/cell types affected by noncoding genetic variants.ResultsGenome‐wide INFERNO analysis identified 168 distinct loci by pruning all genome‐wide significant (p‐value < 5e‐8) GWAS variants using LD structure, a significant increase compared to original top‐SNP analysis (19 loci). Across these 168 distinct loci, INFERNO identified 1,358 candidate potentially causal variants in LD. Around 6% of these variants were localized in coding exons, 67% overlapped enhancers in at least one tissue and 57% overlapped transcription factor binding sites. Co‐localization analysis of GWAS and eQTL data (GTEx_v6p) identified tissue‐specific effects of 53 variants on 43 genes across 16 INFERNO tissue categories, with 12 variants affecting expression of 16 genes in the blood (RP11‐1036E20.9, ZNF223, ZNF285B, RP11‐286H15.1, FAM115D, EPHA1‐AS1, TAS2R60, TAS2R62P, PTK2B) and brain (CR1, ABCA7, POLR2E, EML2, PPM1N, RP11‐385F7.1, FAM115C). These INFERNO‐identified genes include both novel and genes previously implicated in AD. E.g. EPHA1‐AS1 activity was confirmed by luciferase reporter assay [Amlie‐Wolf et al., 2019]. Additionally, our analysis found that PTK2B gene, an early marker of Tau pathology [Dourlen et al., 2016], is regulated by rs28834970 that also disrupts binding motif of HLF(bZIP) in blood.ConclusionGenome‐wide INFERNO analysis confirmed experimentally validated and discovered additional plausible biological mechanisms related to AD. These can help to inform the development of better AD therapeutic targets.

Use of >100,000 NHLBI Trans-Omics for Precision Medicine (TOPMed) Consortium whole genome sequences improves imputation quality and detection of rare variant associations in admixed African and Hispanic/Latino populations.

Most genome-wide association and fine-mapping studies to date have been conducted in individuals of European descent, and genetic studies of populations of Hispanic/Latino and African ancestry are limited. In addition, these populations have more complex linkage disequilibrium structure. In order to better define the genetic architecture of these understudied populations, we leveraged >100,000 phased sequences available from deep-coverage whole genome sequencing through the multi-ethnic NHLBI Trans-Omics for Precision Medicine (TOPMed) program to impute genotypes into admixed African and Hispanic/Latino samples with genome-wide genotyping array data. We demonstrated that using TOPMed sequencing data as the imputation reference panel improves genotype imputation quality in these populations, which subsequently enhanced gene-mapping power for complex traits. For rare variants with minor allele frequency (MAF) < 0.5%, we observed a 2.3- to 6.1-fold increase in the number of well-imputed variants, with 11–34% improvement in average imputation quality, compared to the state-of-the-art 1000 Genomes Project Phase 3 and Haplotype Reference Consortium reference panels. Impressively, even for extremely rare variants with minor allele count <10 (including singletons) in the imputation target samples, average information content rescued was >86%. Subsequent association analyses of TOPMed reference panel-imputed genotype data with hematological traits (hemoglobin (HGB), hematocrit (HCT), and white blood cell count (WBC)) in ~21,600 African-ancestry and ~21,700 Hispanic/Latino individuals identified associations with two rare variants in the HBB gene (rs33930165 with higher WBC [p = 8.8x10-15] in African populations, rs11549407 with lower HGB [p = 1.5x10-12] and HCT [p = 8.8x10-10] in Hispanics/Latinos). By comparison, neither variant would have been genome-wide significant if either 1000 Genomes Project Phase 3 or Haplotype Reference Consortium reference panels had been used for imputation. Our findings highlight the utility of the TOPMed imputation reference panel for identification of novel rare variant associations not previously detected in similarly sized genome-wide studies of under-represented African and Hispanic/Latino populations.

Weighting sequence variants based on their annotation increases the power of genome-wide association studies in dairy cattle

BackgroundGenome-wide association studies (GWAS) are widely used to identify regions of the genome that harbor genetic determinants of quantitative traits. However, the multiple-testing burden from scanning tens of millions of whole-genome sequence variants reduces the power to identify associated variants, especially if sample size is limited. In addition, factors such as inaccuracy of imputation, complex linkage disequilibrium structures, and multiple closely-located causal variants may result in an identified causative mutation not being the most significant single nucleotide polymorphism in a particular genomic region. Therefore, the use of information from different sources, particularly variant annotations, was proposed to enhance the fine-mapping of causal variants. Here, we tested whether applying significance thresholds based on variant annotation categories increases the power of GWAS compared with a flat Bonferroni multiple-testing correction.ResultsWhole-genome sequence variants in dairy cattle were categorized according to type and predicted impact. Then, GWAS between markers and 17 quantitative traits were analyzed for enrichment for association of each annotation category. By using annotation categories that were determined with the variants effect predictor software and datasets indicating regions of open chromatin, “low impact” variants were found to be highly enriched. Moreover, when the variants annotated as “modifier” and not located at open chromatin regions were further classified into different types of potential regulatory elements, the high impact variants, moderate impact variants, variants located in the 3′ and 5′ untranslated regions, and variants located in potential non-coding RNA regions exhibited relatively more enrichment. In contrast, a similar study on human GWAS data reported that enrichment of association signals was highest with high impact variants. We observed an increase in power when these variant category-based significance thresholds were applied for GWAS results on stature in Nordic Holstein cattle, as more candidate genes from previous large GWAS meta-analysis for cattle stature were confirmed.ConclusionsUse of variant category-based genome-wide significance thresholds can marginally increase the power to detect the candidate genes in cattle. With the continued improvements in annotation of the bovine genome, we anticipate that the growing usefulness of variant category-based significance thresholds will be demonstrated.

Genomic differentiation as a tool for single nucleotide polymorphism prioritization for Genome wide association and phenotype prediction in livestock

Abstract Genome-wide association studies (GWAS) have been successful in detecting associations between single nucleotide polymorphisms (SNPs) and phenotypic variation and in identifying several causative mutations. However, SNPs with significant association identified using GWAS tend to explain only small fraction of the phenotypic variations. GWAS are affected by lack of power due to small sample size, large numbers of highly correlated markers, and the moderate to small effects of most quantitative trait loci (QTLs). This situation is further complicated by the continuous increase in marker density, especially with the availability of next-generation sequencing (NGS) data. The latter generates an unprecedented number of marker variants, with a complex linkage disequilibrium (LD) structure limiting the advantage and adequacy of existing methods that internally try to prioritize (filter) SNPs (e.g. BayesB, and BayesR). Consequently, it is becoming necessary to either filter SNPs before conducting the association analysis or to enlist additional sources of information. Methods that include biological prior information (e.g. BayesRC) are limited by the amount and quality of available prior information. Knowledge of genetic diversity based on evolutionary forces is beneficial for tracking loci influenced by selection. The fixation index (F ST ), as a measure of allele frequency variation among sub-populations, provides a tool to reveal genomic regions under selection pressure. In order to evaluate its usefulness as an additional source of information, a simulation was carried out. A trait with heritability of 0.4 was simulated and three subpopulations were created based on the empirical phenotypic distribution ( 95% quantile; and between 5% and 95% quantiles). Marker data was simulated to mimic a bovine chip of 600 K, 1 million, and 3 million SNP marker panels. Genetic complexity of the trait was modelled by the number of QTLs, their distribution, and the magnitude of their effects. Using different empirical cut off values for F ST , most QTLs were correctly detected using as few as 2.5% of SNP markers in the panels. Furthermore, the genomic similarity, calculated based on the selected SNPs, was very high (>0.80) for individuals with similar genetic and phenotypic values despite having limited to no pedigree relationship. These results indicate that filtering SNPs using F ST could be beneficial for use in GWAS by focusing on genome regions under selection pressure. High functional genomic similarity based on selected markers indicates similarity in SNP signatures, regardless of relatedness, and translates into high phenotypic correlation that could be used in decision making.

Assembly and analysis of 100 full MHC haplotypes from the Danish population.

Genes in the major histocompatibility complex (MHC, also known as HLA) play a critical role in the immune response and variation within the extended 4-Mb region shows association with major risks of many diseases. Yet, deciphering the underlying causes of these associations is difficult because the MHC is the most polymorphic region of the genome with a complex linkage disequilibrium structure. Here, we reconstruct full MHC haplotypes from de novo assembled trios without relying on a reference genome and perform evolutionary analyses. We report 100 full MHC haplotypes and call a large set of structural variants in the regions for future use in imputation with GWAS data. We also present the first complete analysis of the recombination landscape in the entire region and show how balancing selection at classical genes have linked effects on the frequency of variants throughout the region.

Reveel: large-scale population genotyping using low-coverage sequencing data.

Population low-coverage whole-genome sequencing is rapidly emerging as a prominent approach for discovering genomic variation and genotyping a cohort. This approach combines substantially lower cost than full-coverage sequencing with whole-genome discovery of low-allele frequency variants, to an extent that is not possible with array genotyping or exome sequencing. However, a challenging computational problem arises of jointly discovering variants and genotyping the entire cohort. Variant discovery and genotyping are relatively straightforward tasks on a single individual that has been sequenced at high coverage, because the inference decomposes into the independent genotyping of each genomic position for which a sufficient number of confidently mapped reads are available. However, in low-coverage population sequencing, the joint inference requires leveraging the complex linkage disequilibrium (LD) patterns in the cohort to compensate for sparse and missing data in each individual. The potentially massive computation time for such inference, as well as the missing data that confound low-frequency allele discovery, need to be overcome for this approach to become practical. Here, we present Reveel, a novel method for single nucleotide variant calling and genotyping of large cohorts that have been sequenced at low coverage. Reveel introduces a novel technique for leveraging LD that deviates from previous Markov-based models, and which is aimed at computational efficiency as well as accuracy in capturing LD patterns present in rare haplotypes. We evaluate Reveel's performance through extensive simulations as well as real data from the 1000 Genomes Project, and show that it achieves higher accuracy in low-frequency allele discovery and substantially lower computation cost than previous state-of-the-art methods. http://reveel.stanford.edu/ : serafim@cs.stanford.edu Supplementary data are available at Bioinformatics online.

Mapping epistatic quantitative trait loci.

BackgroundHow to map quantitative trait loci (QTL) with epistasis efficiently and reliably has been a persistent problem for QTL mapping analysis. There are a number of difficulties for studying epistatic QTL. Linkage can impose a significant challenge for finding epistatic QTL reliably. If multiple QTL are in linkage and have interactions, searching for QTL can become a very delicate issue. A commonly used strategy that performs a two-dimensional genome scan to search for a pair of QTL with epistasis can suffer from low statistical power and also may lead to false identification due to complex linkage disequilibrium and interaction patterns.ResultsTo tackle the problem of complex interaction of multiple QTL with linkage, we developed a three-stage search strategy. In the first stage, main effect QTL are searched and mapped. In the second stage, epistatic QTL that interact significantly with other identified QTL are searched. In the third stage, new epistatic QTL are searched in pairs. This strategy is based on the consideration that most genetic variance is due to the main effects of QTL. Thus by first mapping those main-effect QTL, the statistical power for the second and third stages of analysis for mapping epistatic QTL can be maximized. The search for main effect QTL is robust and does not bias the search for epistatic QTL due to a genetic property associated with the orthogonal genetic model that the additive and additive by additive variances are independent despite of linkage. The model search criterion is empirically and dynamically evaluated by using a score-statistic based resampling procedure. We demonstrate through simulations that the method has good power and low false positive in the identification of QTL and epistasis.ConclusionThis method provides an effective and powerful solution to map multiple QTL with complex epistatic pattern. The method has been implemented in the user-friendly computer software Windows QTL Cartographer. This will greatly facilitate the application of the method for QTL mapping data analysis.Electronic supplementary materialThe online version of this article (doi:10.1186/s12863-014-0112-9) contains supplementary material, which is available to authorized users.

Genome-Wide Association Study of Breast Cancer in the Japanese Population

Breast cancer is the most common malignancy among women in worldwide including Japan. Several studies have identified common genetic variants to be associated with the risk of breast cancer. Due to the complex linkage disequilibrium structure and various environmental exposures in different populations, it is essential to identify variants associated with breast cancer in each population, which subsequently facilitate the better understanding of mammary carcinogenesis. In this study, we conducted a genome-wide association study (GWAS) as well as whole-genome imputation with 2,642 cases and 2,099 unaffected female controls. We further examined 13 suggestive loci (P<1.0×10−5) using an independent sample set of 2,885 cases and 3,395 controls and successfully validated two previously-reported loci, rs2981578 (combined P-value of 1.31×10−12, OR = 1.23; 95% CI = 1.16–.30) on chromosome 10q26 (FGFR2), rs3803662 (combined P-value of 2.79×10−11, OR = 1.21; 95% CI = 1.15–.28) and rs12922061 (combined P-value of 3.97×10−10, OR = 1.23; 95% CI = 1.15–.31) on chromosome 16q12 (TOX3-LOC643714). Weighted genetic risk score on the basis of three significantly associated variants and two previously reported breast cancer associated loci in East Asian population revealed that individuals who carry the most risk alleles in category 5 have 2.2 times higher risk of developing breast cancer in the Japanese population than those who carry the least risk alleles in reference category 1. Although we could not identify additional loci associated with breast cancer, our study utilized one of the largest sample sizes reported to date, and provided genetic status that represent the Japanese population. Further local and international collaborative study is essential to identify additional genetic variants that could lead to a better, accurate prediction for breast cancer.

GUESS-ing Polygenic Associations with Multiple Phenotypes Using a GPU-Based Evolutionary Stochastic Search Algorithm

Genome-wide association studies (GWAS) yielded significant advances in defining the genetic architecture of complex traits and disease. Still, a major hurdle of GWAS is narrowing down multiple genetic associations to a few causal variants for functional studies. This becomes critical in multi-phenotype GWAS where detection and interpretability of complex SNP(s)-trait(s) associations are complicated by complex Linkage Disequilibrium patterns between SNPs and correlation between traits. Here we propose a computationally efficient algorithm (GUESS) to explore complex genetic-association models and maximize genetic variant detection. We integrated our algorithm with a new Bayesian strategy for multi-phenotype analysis to identify the specific contribution of each SNP to different trait combinations and study genetic regulation of lipid metabolism in the Gutenberg Health Study (GHS). Despite the relatively small size of GHS (n = 3,175), when compared with the largest published meta-GWAS (n>100,000), GUESS recovered most of the major associations and was better at refining multi-trait associations than alternative methods. Amongst the new findings provided by GUESS, we revealed a strong association of SORT1 with TG-APOB and LIPC with TG-HDL phenotypic groups, which were overlooked in the larger meta-GWAS and not revealed by competing approaches, associations that we replicated in two independent cohorts. Moreover, we demonstrated the increased power of GUESS over alternative multi-phenotype approaches, both Bayesian and non-Bayesian, in a simulation study that mimics real-case scenarios. We showed that our parallel implementation based on Graphics Processing Units outperforms alternative multi-phenotype methods. Beyond multivariate modelling of multi-phenotypes, our Bayesian model employs a flexible hierarchical prior structure for genetic effects that adapts to any correlation structure of the predictors and increases the power to identify associated variants. This provides a powerful tool for the analysis of diverse genomic features, for instance including gene expression and exome sequencing data, where complex dependencies are present in the predictor space.

Genetic Variation in the Extended Major Histocompatibility Complex and Susceptibility to Childhood Acute Lymphoblastic Leukemia: A Review of the Evidence

The enduring suspicion that infections and immunologic response may play a role in the etiology of childhood leukemia, particularly acute lymphoblastic leukemia (ALL), is now supported, albeit still indirectly, by numerous epidemiological studies. The cumulative evidence includes, for example, descriptive observations of a peculiar peak incidence at age 2–5 years for ALL in economically developed countries, clustering of cases in situations of population mixing associated with unusual patterns of personal contacts, associations with various proxy measures for immune modulatory exposures early in life, and genetic susceptibility conferred by variation in genes involved in the immune system. In this review, our focus is the extended major histocompatibility complex (MHC), an approximately 7.6 Mb region that is well-known for its high-density of expressed genes, extensive polymorphisms exhibiting complex linkage disequilibrium patterns, and its disproportionately large number of immune-related genes, including human leukocyte antigen (HLA). First discovered through the role they play in transplant rejection, the classical HLA class I (HLA-A, -B, and -C) and class II (HLA-DR, HLA-DQ, and HLA-DP) molecules reside at the epicenter of the immune response pathways and are now the targets of many disease susceptibility studies, including those for childhood leukemia. The genes encoding the HLA molecules are only a minority of the over 250 expressed genes in the xMHC, and a growing number of studies are beginning to evaluate other loci through targeted investigations or utilizing a mapping approach with a comprehensive screen of the entire region. Here, we review the current epidemiologic evidence available to date regarding genetic variation contained within this highly unique region of the genome and its relationship with childhood ALL risk.