Strategies for phasing and imputation in a population isolate

Genetic architecture distinguishes tinnitus from hearing loss

Tinnitus is a heritable, highly prevalent auditory disorder treated by multiple medical specialties. Previous GWAS indicated high genetic correlations between tinnitus and hearing loss, with little indication of differentiating signals. We present a GWAS meta-analysis, triple previous sample sizes, and expand to non-European ancestries. GWAS in 596,905 Million Veteran Program subjects identified 39 tinnitus loci, and identified genes related to neuronal synapses and cochlear structural support. Applying state-of-the-art analytic tools, we confirm a large number of shared variants, but also a distinct genetic architecture of tinnitus, with higher polygenicity and large proportion of variants not shared with hearing difficulty. Tissue-expression analysis for tinnitus infers broad enrichment across most brain tissues, in contrast to hearing difficulty. Finally, tinnitus is not only correlated with hearing loss, but also with a spectrum of psychiatric disorders, providing potential new avenues for treatment. This study establishes tinnitus as a distinct disorder separate from hearing difficulties.

How local reference panels improve imputation in French populations

Imputation servers offer the exclusive possibility to harness the largest public reference panels which have been shown to deliver very high precision in the imputation of European genomes. Many studies have nonetheless stressed the importance of 'study specific panels' (SSPs) as an alternative and have shown the benefits of combining public reference panels with SSPs. But such combined approaches are not attainable when using external imputation servers. To investigate how to confront this challenge, we imputed 550 French individuals using either the University of Michigan imputation server with the Haplotype Reference Consortium (HRC) panel or an in-house SSP of 850 whole-genome sequenced French individuals. With approximate geo-localization of both our target and SSP individuals we are able to pinpoint different scenarios where SSP-based imputation would be preferred over server-based imputation or vice-versa. This is achieved by showing to a high degree of resolution the importance of the proximity of the reference panel to target individuals; with a focus on the clear added value of SSPs for estimating haplotype phase and for the imputation of rare variants (minor allele-frequency below 0.01). Such benefits were most evident for individuals from the same geographical regions in France as the SSP individuals. Overall, only 42.3% of all 125,442 variants evaluated were better imputed with an SSP from France compared to an external reference panel, however this rises to 58.1% for individuals from geographic regions well covered by the SSP. By investigating haplotype sharing and population fine-structure in France, we show the importance of including SSP haplotypes for imputation but also that they should ideally be combined with large public panels. In the absence of the unattainable results from a combined panel of the HRC and our French SSP, we put forward a pragmatic solution where server-based and SSP-based imputation outcomes can be combined based on comparing posterior genotype probabilities. We show that such an approach can give a level of imputation accuracy in excess of what could be achieved with either strategy alone. The results presented provide detailed insights into the accuracy of imputation that should be expected from different strategies for European populations.

Genotype-Phenotype Models Predicting V̇O 2max Response to High-Intensity Interval Training in Physically Inactive Chinese

PURPOSE

This study aimed to analyze the interindividual differences of the maximal oxygen uptake (V̇O 2max ) response to 12 wk of high-intensity interval training (HIIT), and the genotype-phenotype models were constructed to predict the effect of HIIT on V̇O 2max .

METHODS

A total of 228 physically inactive adults who completed a 12-wk HIIT were analyzed. A genome-wide association study (GWAS) was conducted to identify genetic variants associated with the V̇O 2max response. Nonresponders, responders, and the highest training responders were defined as the effect sizes (ES) <0.2, ≥0.2, and ≥0.8, respectively. We generated polygenic predictor score (PPS) using lead variants and constructed a predictive model for V̇O 2max response based on a linear stepwise regression analysis.

RESULTS

The V̇O 2max increased significantly after HIIT (~14%, P < 0.001), but with interindividual differences (-7.8 to 17.9 mL·kg -1 ·min -1 ). In 27% of participants, the V̇O 2max showed no improvement. We identified one genetic locus near the γ-aminobutyric acid type A receptor subunit beta 3 gene ( GABRB3 , rs17116985) associated with V̇O 2max response at the genome-wide significance level ( P < 5 × 10 -8 ), and an additional nine single nucleotide polymorphisms (SNPs) at the suggestive significance level ( P < 1 × 10 -5 ). The SNPs rs474377, rs9365605, and rs17116985, respectively, explained 11%, 9%, and 6.2% of variance in V̇O 2max response. The 13 SNPs ( P < 1 × 10 -5 ) were found on chromosome 6 (position: 148209316-148223568). Individuals with a PPS greater than 1.757 had the highest response, and those with a PPS lower than -3.712 were nonresponders. The PPS, baseline V̇O 2max , sex, and body mass explained 56.4% of the variance in the V̇O 2max response; the major predictor was the PPS, which explained 39.4% of the variance.

CONCLUSIONS

The PPS, baseline V̇O 2max , sex, and body mass could explain the variance in V̇O 2max response. Individuals who had a PPS greater than 1.757 had the highest training response after 12 wk of HIIT. Genetic variants in a region on chromosome 6, especially the sterile alpha motif domain containing 5 gene ( SAMD5 ), which had been explored influencing angiogenesis, might have a potential role in the V̇O 2max response.

Opening the Black Box of Imputation Software to Study the Impact of Reference Panel Composition on Performance

Genotype imputation is widely used to enrich genetic datasets. The operation relies on panels of known reference haplotypes, typically with whole-genome sequencing data. How to choose a reference panel has been widely studied and it is essential to have a panel that is well matched to the individuals who require missing genotype imputation. However, it is broadly accepted that such an imputation panel will have an enhanced performance with the inclusion of diversity (haplotypes from many different populations). We investigate this observation by examining, in fine detail, exactly which reference haplotypes are contributing at different regions of the genome. This is achieved using a novel method of inserting synthetic genetic variation into the reference panel in order to track the performance of leading imputation algorithms. We show that while diversity may globally improve imputation accuracy, there can be occasions where incorrect genotypes are imputed following the inclusion of more diverse haplotypes in the reference panel. We, however, demonstrate a technique for retaining and benefitting from the diversity in the reference panel whilst avoiding the occasional adverse effects on imputation accuracy. What is more, our results more clearly elucidate the role of diversity in a reference panel than has been shown in previous studies.

A Pipeline for Phasing and Genotype Imputation on Mixed Human Data (Parents-Offspring Trios and Unrelated Subjects) by Reviewing Current Methods and Software

Genotype imputation has become an essential prerequisite when performing association analysis. It is a computational technique that allows us to infer genetic markers that have not been directly genotyped, thereby increasing statistical power in subsequent association studies, which consequently has a crucial impact on the identification of causal variants. Many features need to be considered when choosing the proper algorithm for imputation, including the target sample on which it is performed, i.e., related individuals, unrelated individuals, or both. Problems could arise when dealing with a target sample made up of mixed data, composed of both related and unrelated individuals, especially since the scientific literature on this topic is not sufficiently clear. To shed light on this issue, we examined existing algorithms and software for performing phasing and imputation on mixed human data from SNP arrays, specifically when related subjects belong to trios. By discussing the advantages and limitations of the current algorithms, we identified LD-based methods as being the most suitable for reconstruction of haplotypes in this specific context, and we proposed a feasible pipeline that can be used for imputing genotypes in both phased and unphased human data.

A comparative analysis of current phasing and imputation software

Whole-genome data has become significantly more accessible over the last two decades. This can largely be attributed to both reduced sequencing costs and imputation models which make it possible to obtain nearly whole-genome data from less expensive genotyping methods, such as microarray chips. Although there are many different approaches to imputation, the Hidden Markov Model (HMM) remains the most widely used. In this study, we compared the latest versions of the most popular HMM-based tools for phasing and imputation: Beagle5.4, Eagle2.4.1, Shapeit4, Impute5 and Minimac4. We benchmarked them on four input datasets with three levels of chip density. We assessed each imputation software on the basis of accuracy, speed and memory usage, and showed how the choice of imputation accuracy metric can result in different interpretations. The highest average concordance rate was achieved by Beagle5.4, followed by Impute5 and Minimac4, using a reference-based approach during phasing and the highest density chip. IQS and R2 metrics revealed that Impute5 and Minimac4 obtained better results for low frequency markers, while Beagle5.4 remained more accurate for common markers (MAF>5%). Computational load as measured by run time was lower for Beagle5.4 than Minimac4 and Impute5, while Minimac4 utilized the least memory of the imputation tools we compared. ShapeIT4, used the least memory of the phasing tools examined with genotype chip data, while Eagle2.4.1 used the least memory phasing WGS data. Finally, we determined the combination of phasing software, imputation software, and reference panel, best suited for different situations and analysis needs and created an automated pipeline that provides a way for users to create customized chips designed to optimize their imputation results.

Moment estimators of relatedness from low-depth whole-genome sequencing data

Background

Estimating relatedness is an important step for many genetic study designs. A variety of methods for estimating coefficients of pairwise relatedness from genotype data have been proposed. Both the kinship coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varphi$$\end{document}φ and the fraternity coefficient \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\psi$$\end{document}ψ for all pairs of individuals are of interest. However, when dealing with low-depth sequencing or imputation data, individual level genotypes cannot be confidently called. To ignore such uncertainty is known to result in biased estimates. Accordingly, methods have recently been developed to estimate kinship from uncertain genotypes.

Results

We present new method-of-moment estimators of both the coefficients \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varphi$$\end{document}φ and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\psi$$\end{document}ψ calculated directly from genotype likelihoods. We have simulated low-depth genetic data for a sample of individuals with extensive relatedness by using the complex pedigree of the known genetic isolates of Cilento in South Italy. Through this simulation, we explore the behaviour of our estimators, demonstrate their properties, and show advantages over alternative methods. A demonstration of our method is given for a sample of 150 French individuals with down-sampled sequencing data.

Conclusions

We find that our method can provide accurate relatedness estimates whilst holding advantages over existing methods in terms of robustness, independence from external software, and required computation time. The method presented in this paper is referred to as LowKi (Low-depth Kinship) and has been made available in an R package (https://github.com/genostats/LowKi).

Supplementary Information

The online version contains supplementary material available at 10.1186/s12859-022-04795-8.

Hearing loss and tinnitus: association studies for complex-hearing disorders in mouse and man

Genome-wide association studies (GWAS) provide an unbiased first look at genetic loci involved in aging and noise-induced sensorineural hearing loss and tinnitus. The hearing phenotype, whether audiogram-based or self-report, is regressed against genotyped information at representative single nucleotide polymorphisms (SNPs) across the genome. Findings include the fact that both hearing loss and tinnitus are polygenic disorders, with up to thousands of genes, each of effect size of < 0.02. Smaller human GWAS' were able to use objective measures and identified a few loci; however, hundreds of thousands of participants have been required for the statistical power to identify significant variants, and GWAS is unable to assess rare variants with mean allele frequency < 1%. Animal studies are required as well because of inability to access the human cochlea. Mouse GWAS builds on linkage techniques and the known phenotypic differences in auditory function between inbred strains. With the advantage that the laboratory environment can be controlled for noise and aging, the Hybrid Mouse Diversity Panel (HDMP) combines 100 strains sequenced at high resolution. Lift-over regions between mice and humans have identified over 17,000 homologous genes. Since most significant SNPs are either intergenic or in introns, and binding sites between species are poorly preserved between species, expression quantitative trait locus information is required to bring humans and mice into agreement. Transcriptome-wide analysis studies (TWAS) can prioritize putative causal genes and tissues. Diverse species, each making a distinct contribution, carry a synergistic advantage in the quest for treatment and ultimate cure of sensorineural hearing difficulties.

Very low-depth whole-genome sequencing in complex trait association studies

Motivation

Very low-depth sequencing has been proposed as a cost-effective approach to capture low-frequency and rare variation in complex trait association studies. However, a full characterization of the genotype quality and association power for very low-depth sequencing designs is still lacking.

Results

We perform cohort-wide whole-genome sequencing (WGS) at low depth in 1239 individuals (990 at 1× depth and 249 at 4× depth) from an isolated population, and establish a robust pipeline for calling and imputing very low-depth WGS genotypes from standard bioinformatics tools. Using genotyping chip, whole-exome sequencing (75× depth) and high-depth (22×) WGS data in the same samples, we examine in detail the sensitivity of this approach, and show that imputed 1× WGS recapitulates 95.2% of variants found by imputed GWAS with an average minor allele concordance of 97% for common and low-frequency variants. In our study, 1× further allowed the discovery of 140 844 true low-frequency variants with 73% genotype concordance when compared to high-depth WGS data. Finally, using association results for 57 quantitative traits, we show that very low-depth WGS is an efficient alternative to imputed GWAS chip designs, allowing the discovery of up to twice as many true association signals than the classical imputed GWAS design.

Availability and implementation

The HELIC genotype and WGS datasets have been deposited to the European Genome-phenome Archive (https://www.ebi.ac.uk/ega/home): EGAD00010000518; EGAD00010000522; EGAD00010000610; EGAD00001001636, EGAD00001001637. The peakplotter software is available at https://github.com/wtsi-team144/peakplotter, the transformPhenotype app can be downloaded at https://github.com/wtsi-team144/transformPhenotype.

Supplementary information

Supplementary data are available at Bioinformatics online.

Kinpute: using identity by descent to improve genotype imputation

MOTIVATION

Genotype imputation, though generally accurate, often results in many genotypes being poorly imputed, particularly in studies where the individuals are not well represented by standard reference panels. When individuals in the study share regions of the genome identical by descent (IBD), it is possible to use this information in combination with a study-specific reference panel (SSRP) to improve the imputation results. Kinpute uses IBD information-due to recent, familial relatedness or distant, unknown ancestors-in conjunction with the output from linkage disequilibrium (LD) based imputation methods to compute more accurate genotype probabilities. Kinpute uses a novel method for IBD imputation, which works even in the absence of a pedigree, and results in substantially improved imputation quality.

RESULTS

Given initial estimates of average IBD between subjects in the study sample, Kinpute uses a novel algorithm to select an optimal set of individuals to sequence and use as an SSRP. Kinpute is designed to use as input both this SSRP and the genotype probabilities output from other LD-based imputation software, and uses a new method to combine the LD imputed genotype probabilities with IBD configurations to substantially improve imputation. We tested Kinpute on a human population isolate where 98 individuals have been sequenced. In half of this sample, whose sequence data was masked, we used Impute2 to perform LD-based imputation and Kinpute was used to obtain higher accuracy genotype probabilities. Measures of imputation accuracy improved significantly, particularly for those genotypes that Impute2 imputed with low certainty.

AVAILABILITY AND IMPLEMENTATION

Kinpute is an open-source and freely available C++ software package that can be downloaded from.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Exploring effective approaches for haplotype block phasing

BACKGROUND

Knowledge of phase, the specific allele sequence on each copy of homologous chromosomes, is increasingly recognized as critical for detecting certain classes of disease-associated mutations. One approach for detecting such mutations is through phased haplotype association analysis. While the accuracy of methods for phasing genotype data has been widely explored, there has been little attention given to phasing accuracy at haplotype block scale. Understanding the combined impact of the accuracy of phasing tool and the method used to determine haplotype blocks on the error rate within the determined blocks is essential to conduct accurate haplotype analyses.

RESULTS

We present a systematic study exploring the relationship between seven widely used phasing methods and two common methods for determining haplotype blocks. The evaluation focuses on the number of haplotype blocks that are incorrectly phased. Insights from these results are used to develop a haplotype estimator based on a consensus of three tools. The consensus estimator achieved the most accurate phasing in all applied tests. Individually, EAGLE2, BEAGLE and SHAPEIT2 alternate in being the best performing tool in different scenarios. Determining haplotype blocks based on linkage disequilibrium leads to more correctly phased blocks compared to a sliding window approach. We find that there is little difference between phasing sections of a genome (e.g. a gene) compared to phasing entire chromosomes. Finally, we show that the location of phasing error varies when the tools are applied to the same data several times, a finding that could be important for downstream analyses.

CONCLUSIONS

The choice of phasing and block determination algorithms and their interaction impacts the accuracy of phased haplotype blocks. This work provides guidance and evidence for the different design choices needed for analyses using haplotype blocks. The study highlights a number of issues that may have limited the replicability of previous haplotype analysis.

Rare Variants Imputation in Admixed Populations: Comparison Across Reference Panels and Bioinformatics Tools

Background

Imputation has become a standard approach in genome-wide association studies (GWAS) to infer in silico untyped markers. Although feasibility for common variants imputation is well established, we aimed to assess rare and ultra-rare variants’ imputation in an admixed Caribbean Hispanic population (CH).

Methods

We evaluated imputation accuracy in CH (N = 1,000), focusing on rare (0.1% ≤ minor allele frequency (MAF) ≤ 1%) and ultra-rare (MAF < 0.1%) variants. We used two reference panels, the Haplotype Reference Consortium (HRC; N = 27,165) and 1000 Genome Project (1000G phase 3; N = 2,504) and multiple phasing (SHAPEIT, Eagle2) and imputation algorithms (IMPUTE2, MACH-Admix). To assess imputation quality, we reported: (a) high-quality variant counts according to imputation tools’ internal indexes (e.g., IMPUTE2 “Info” ≥ 80%). (b) Wilcoxon Signed-Rank Test comparing imputation quality for genotyped variants that were masked and imputed; (c) Cohen’s kappa coefficient to test agreement between imputed and whole-exome sequencing (WES) variants; (d) imputation of G206A mutation in the PSEN1 (ultra-rare in the general population an more frequent in CH) followed by confirmation genotyping. We also tested ancestry proportion (European, African and Native American) against WES-imputation mismatches in a Poisson regression fashion.

Results

SHAPEIT2 retrieved higher percentage of imputed high-quality variants than Eagle2 (rare: 51.02% vs. 48.60%; ultra-rare 0.66% vs. 0.65%, Wilcoxon p-value < 0.001). SHAPEIT-IMPUTE2 employing HRC outperformed 1000G (64.50% vs. 59.17%; 1.69% vs. 0.75% for high-quality rare and ultra-rare variants, respectively, Wilcoxon p-value < 0.001). SHAPEIT-IMPUTE2 outperformed MaCH-Admix. Compared to 1000G, HRC-imputation retrieved a higher number of high-quality rare and ultra-rare variants, despite showing lower agreement between imputed and WES variants (e.g., rare: 98.86% for HRC vs. 99.02% for 1000G). High Kappa (K = 0.99) was observed for both reference panels. Twelve G206A mutation carriers were imputed and all validated by confirmation genotyping. African ancestry was associated with higher imputation errors for uncommon and rare variants (p-value < 1e-05).

Conclusion

Reference panels with larger numbers of haplotypes can improve imputation quality for rare and ultra-rare variants in admixed populations such as CH. Ethnic composition is an important predictor of imputation accuracy, with higher African ancestry associated with poorer imputation accuracy.

Whole-Exome Sequencing in the Isolated Populations of Cilento from South Italy

The present study describes the genetic architecture of the isolated populations of Cilento, through the analysis of exome sequence data of 245 representative individuals of these populations. By annotating the exome variants and cataloguing them according to their frequency and functional effects, we identified 347,684 variants, 67.4% of which are rare and low frequency variants, and 1% of them (corresponding to 319 variants per person) are classified as high functional impact variants; also, 39,946 (11.5% of the total) are novel variants, for which we determined a significant enrichment for deleterious effects. By comparing the allele frequencies in Cilento with those from the Tuscan population from the 1000 Genomes Project Phase 3, we highlighted an increase in allele frequency in Cilento especially for variants which map to genes involved in extracellular matrix formation and organization. Furthermore, among the variants showing increased frequency we identified several known rare disease-causing variants. By different population genetics analyses, we corroborated the status of the Cilento populations as genetic isolates. Finally, we showed that exome data of Cilento represents a useful local reference panel capable of improving the accuracy of genetic imputation, thus adding power to genetic studies of human traits in these populations.

Detecting the dominance component of heritability in isolated and outbred human populations

Inconsistencies between published estimates of dominance heritability between studies of human genetic isolates and human outbred populations incite investigation into whether such differences result from particular trait architectures or specific population structures. We analyse simulated datasets, characteristic of genetic isolates and of unrelated individuals, before analysing the isolate of Cilento for various commonly studied traits. We show the strengths of using genetic relationship matrices for variance decomposition over identity-by-descent based methods in a population isolate and that heritability estimates in isolates will avoid the downward biases that may occur in studies of samples of unrelated individuals; irrespective of the simulated distribution of causal variants. Yet, we also show that precise estimates of dominance in isolates are demonstrably problematic in the presence of shared environmental effects and such effects should be accounted for. Nevertheless, we demonstrate how studying isolates can help determine the existence or non-existence of dominance for complex traits, and we find strong indications of non-zero dominance for low-density lipoprotein level in Cilento. Finally, we recommend future study designs to analyse trait variance decomposition from ensemble data across multiple population isolates.

Comparison and assessment of family- and population-based genotype imputation methods in large pedigrees

Genotype imputation is widely used in genome-wide association studies to boost variant density, allowing increased power in association testing. Many studies currently include pedigree data due to increasing interest in rare variants coupled with the availability of appropriate analysis tools. The performance of population-based (subjects are unrelated) imputation methods is well established. However, the performance of family- and population-based imputation methods on family data has been subject to much less scrutiny. Here, we extensively compare several family- and population-based imputation methods on family data of large pedigrees with both European and African ancestry. Our comparison includes many widely used family- and population-based tools and another method, Ped_Pop, which combines family- and population-based imputation results. We also compare four subject selection strategies for full sequencing to serve as the reference panel for imputation: GIGI-Pick, ExomePicks, PRIMUS, and random selection. Moreover, we compare two imputation accuracy metrics: the Imputation Quality Score and Pearson's correlation R ² for predicting power of association analysis using imputation results. Our results show that (1) GIGI outperforms Merlin; (2) family-based imputation outperforms population-based imputation for rare variants but not for common ones; (3) combining family- and population-based imputation outperforms all imputation approaches for all minor allele frequencies; (4) GIGI-Pick gives the best selection strategy based on the R ² criterion; and (5) R ² is the best measure of imputation accuracy. Our study is the first to extensively evaluate the imputation performance of many available family- and population-based tools on the same family data and provides guidelines for future studies.