Technology-specific error signatures in the 1000 Genomes Project data

Next-generation sequencing (NGS) will likely facilitate a better understanding of the causes and consequences of human genetic variability. In this context, the validity of NGS-inferred single-nucleotide variants (SNVs) is of paramount importance. We therefore developed a statistical framework to assess the fidelity of three common NGS platforms. Using aligned DNA sequence data from two completely sequenced HapMap samples as included in the 1000 Genomes Project, we unraveled remarkably different error profiles for the three platforms. Compared to confirmed HapMap variants, newly identified SNVs included a substantial proportion of false positives (3–17%). Consensus calling by more than one platform yielded significantly lower error rates (1–4%). This implies that the use of multiple NGS platforms may be more cost-efficient than relying upon a single technology alone, particularly in physically localized sequencing experiments that rely upon small error rates. Our study thus highlights that different NGS platforms suit different practical applications differently well, and that NGS-based studies require stringent data quality control for their results to be valid.     

The 1000 Genomes Project: data management and community access

The 1000 Genomes Project was launched as one of the largest distributed data collection and analysis projects ever undertaken in biology. In addition to the primary scientific goals of creating both a deep catalog of human genetic variation and extensive methods to accurately discover and characterize variation using new sequencing technologies, the project makes all of its data publicly available. Members of the project data coordination center have developed and deployed several tools to enable widespread data access.     

Performance of Genotype Imputation for Low Frequency and Rare Variants from the 1000 Genomes

Genotype imputation is now routinely applied in genome-wide association studies (GWAS) and meta-analyses. However, most of the imputations have been run using HapMap samples as reference, imputation of low frequency and rare variants (minor allele frequency (MAF) < 5%) are not systemically assessed. With the emergence of next-generation sequencing, large reference panels (such as the 1000 Genomes panel) are available to facilitate imputation of these variants. Therefore, in order to estimate the performance of low frequency and rare variants imputation, we imputed 153 individuals, each of whom had 3 different genotype array data including 317k, 610k and 1 million SNPs, to three different reference panels: the 1000 Genomes pilot March 2010 release (1KGpilot), the 1000 Genomes interim August 2010 release (1KGinterim), and the 1000 Genomes phase1 November 2010 and May 2011 release (1KGphase1) by using IMPUTE version 2. The differences between these three releases of the 1000 Genomes data are the sample size, ancestry diversity, number of variants and their frequency spectrum. We found that both reference panel and GWAS chip density affect the imputation of low frequency and rare variants. 1KGphase1 outperformed the other 2 panels, at higher concordance rate, higher proportion of well-imputed variants (info>0.4) and higher mean info score in each MAF bin. Similarly, 1M chip array outperformed 610K and 317K. However for very rare variants (MAF≤0.3%), only 0–1% of the variants were well imputed. We conclude that the imputation of low frequency and rare variants improves with larger reference panels and higher density of genome-wide genotyping arrays. Yet, despite a large reference panel size and dense genotyping density, very rare variants remain difficult to impute.     

High-coverage whole-genome sequencing of the expanded 1000 Genomes Project cohort including 602 trios

1. Download : Download high-res image (257KB)
2. Download : Download full-size image

     

Amino Acid Changes in Disease-Associated Variants Differ Radically from Variants Observed in the 1000 Genomes Project Dataset

The 1000 Genomes Project data provides a natural background dataset for amino acid germline mutations in humans. Since the direction of mutation is known, the amino acid exchange matrix generated from the observed nucleotide variants is asymmetric and the mutabilities of the different amino acids are very different. These differences predominantly reflect preferences for nucleotide mutations in the DNA (especially the high mutation rate of the CpG dinucleotide, which makes arginine mutability very much higher than other amino acids) rather than selection imposed by protein structure constraints, although there is evidence for the latter as well. The variants occur predominantly on the surface of proteins (82%), with a slight preference for sites which are more exposed and less well conserved than random. Mutations to functional residues occur about half as often as expected by chance. The disease-associated amino acid variant distributions in OMIM are radically different from those expected on the basis of the 1000 Genomes dataset. The disease-associated variants preferentially occur in more conserved sites, compared to 1000 Genomes mutations. Many of the amino acid exchange profiles appear to exhibit an anti-correlation, with common exchanges in one dataset being rare in the other. Disease-associated variants exhibit more extreme differences in amino acid size and hydrophobicity. More modelling of the mutational processes at the nucleotide level is needed, but these observations should contribute to an improved prediction of the effects of specific variants in humans.     

Assessment of Genotype Imputation Performance Using 1000 Genomes in African American Studies

Genotype imputation, used in genome-wide association studies to expand coverage of single nucleotide polymorphisms (SNPs), has performed poorly in African Americans compared to less admixed populations. Overall, imputation has typically relied on HapMap reference haplotype panels from Africans (YRI), European Americans (CEU), and Asians (CHB/JPT). The 1000 Genomes project offers a wider range of reference populations, such as African Americans (ASW), but their imputation performance has had limited evaluation. Using 595 African Americans genotyped on Illumina’s HumanHap550v3 BeadChip, we compared imputation results from four software programs (IMPUTE2, BEAGLE, MaCH, and MaCH-Admix) and three reference panels consisting of different combinations of 1000 Genomes populations (February 2012 release): (1) 3 specifically selected populations (YRI, CEU, and ASW); (2) 8 populations of diverse African (AFR) or European (AFR) descent; and (3) all 14 available populations (ALL). Based on chromosome 22, we calculated three performance metrics: (1) concordance (percentage of masked genotyped SNPs with imputed and true genotype agreement); (2) imputation quality score (IQS; concordance adjusted for chance agreement, which is particularly informative for low minor allele frequency [MAF] SNPs); and (3) average r2hat (estimated correlation between the imputed and true genotypes, for all imputed SNPs). Across the reference panels, IMPUTE2 and MaCH had the highest concordance (91%–93%), but IMPUTE2 had the highest IQS (81%–83%) and average r2hat (0.68 using YRI+ASW+CEU, 0.62 using AFR+EUR, and 0.55 using ALL). Imputation quality for most programs was reduced by the addition of more distantly related reference populations, due entirely to the introduction of low frequency SNPs (MAF≤2%) that are monomorphic in the more closely related panels. While imputation was optimized by using IMPUTE2 with reference to the ALL panel (average r2hat = 0.86 for SNPs with MAF>2%), use of the ALL panel for African American studies requires careful interpretation of the population specificity and imputation quality of low frequency SNPs.     

Haplotype reconstruction from genotype data using Imperfect Phylogeny

Critical to the understanding of the genetic basis for complex diseases is the modeling of human variation. Most of this variation can be characterized by single nucleotide polymorphisms (SNPs) which are mutations at a single nucleotide position. To characterize the genetic variation between different people, we must determine an individual's haplotype or which nucleotide base occurs at each position of these common SNPs for each chromosome. In this paper, we present results for a highly accurate method for haplotype resolution from genotype data. Our method leverages a new insight into the underlying structure of haplotypes that shows that SNPs are organized in highly correlated 'blocks'. In a few recent studies, considerable parts of the human genome were partitioned into blocks, such that the majority of the sequenced genotypes have one of about four common haplotypes in each block. Our method partitions the SNPs into blocks, and for each block, we predict the common haplotypes and each individual's haplotype. We evaluate our method over biological data. Our method predicts the common haplotypes perfectly and has a very low error rate (<2% over the data) when taking into account the predictions for the uncommon haplotypes. Our method is extremely efficient compared with previous methods such as PHASE and HAPLOTYPER. Its efficiency allows us to find the block partition of the haplotypes, to cope with missing data and to work with large datasets. AVAILABILITY: The algorithm is available via a Web server at http://www.calit2.net/compbio/hap/     

Evaluation of microsatellite variation in the 1000 Genomes Project pilot studies is indicative of the quality and utility of the raw data and alignments

We performed an analysis of global microsatellite variation on the two kindreds sequenced at high depth (~20×-60×) in the 1000 Genomes Project pilot studies because alterations in these highly mutable repetitive sequences have been linked with many phenotypes and disease risks. The standard alignment technique performs poorly in microsatellite regions as a consequence of low effective coverage (~1×-5×) resulting in 79% of the informative loci exhibiting non-Mendelian inheritance patterns. We used a more stringent approach in computing robust allelotypes resulting in 94.4% of the 1095 informative repeats conforming to traditional inheritance. The high-confidence allelotypes were analyzed to obtain an estimate of the minimum polymorphism rate as a function of motif length, motif sequence, and distribution within the genome.     

Evaluating the Coverage and Potential of Imputing the Exome Microarray with Next-Generation Imputation Using the 1000 Genomes Project

Next-generation genotyping microarrays have been designed with insights from large-scale sequencing of exomes and whole genomes. The exome genotyping arrays promise to query the functional regions of the human genome at a fraction of the sequencing cost, thus allowing large number of samples to be genotyped. However, two pertinent questions exist: firstly, how representative is the content of the exome chip for populations not involved in the design of the chip; secondly, can the content of the exome chip be imputed with the reference data from the 1000 Genomes Project (1KGP). By deep whole-genome sequencing two Asian populations that are not part of the 1KGP, comprising 96 Southeast Asian Malays and 36 South Asian Indians for which the same samples have also been genotyped on both the Illumina 2.5 M and exome microarrays, we discovered the exome chip is a poor representation of exonic content in our two populations. However, up to 94.1% of the variants on the exome chip that are polymorphic in our populations can be confidently imputed with existing non-exome-centric microarrays using the 1KGP panel. The coverage further increases if there exists population-specific reference data from whole-genome sequencing. There is thus limited gain in using the exome chip for populations not involved in the microarray design. Instead, for the same cost of genotyping 2,000 samples on the exome chip, performing whole-genome sequencing of at least 35 samples in that population to complement the 1KGP may yield a higher coverage of the exonic content from imputation instead.     

Nation-wide litter fall data from 21 forests of the Monitoring Sites 1000 Project in Japan

This data paper reports litter fall data collected in a network of 21 forest sites in Japan. This is the largest litter fall data set freely available in Japan to date. The network is a part of the Monitoring Sites 1000 Project launched by the Ministry of the Environment, Japan. It covers subarctic to subtropical climate zones and the four major forest types in Japan. Twenty-three permanent plots in which usually 25 litter traps were installed were established in old-growth or secondary natural forests. Litter falls were collected monthly from 2004, and sorted into leaves, branches, reproductive structures and miscellaneous. The data provide seasonal patterns and inter-annual dynamics of litter falls, and their geographical patterns, and offer good opportunities for meta-analyses and comparative studies among forests.     

Estimating selfing rates from reconstructed pedigrees using multilocus genotype data

Several methods have been developed to estimate the selfing rate of a population from a sample of individuals genotyped for several marker loci. These methods can be based on homozygosity excess (or inbreeding), identity disequilibrium, progeny array (PA) segregation or population assignment incorporating partial selfing. Progeny array-based method is generally the best because it is not subject to some assumptions made by other methods (such as lack of misgenotyping, absence of biparental inbreeding and presence of inbreeding equilibrium), and it can reveal other facets of a mixed-mating system such as patterns of shared paternity. However, in practice, it is often difficult to obtain PAs, especially for animal species. In this study, we propose a method to reconstruct the pedigree of a sample of individuals taken from a monoecious diploid population practicing mixed mating, using multilocus genotypic data. Selfing and outcrossing events are then detected when an individual derives from identical parents and from two distinct parents, respectively. Selfing rate is estimated by the proportion of selfed offspring in the reconstructed pedigree of a sample of individuals. The method enjoys many advantages of the PA method, but without the need of a priori family structure, although such information, if available, can be utilized to improve the inference. Furthermore, the new method accommodates genotyping errors, estimates allele frequencies jointly and is robust to the presence of biparental inbreeding and inbreeding disequilibrium. Both simulated and empirical data were analysed by the new and previous methods to compare their statistical properties and accuracies.     

Identification of Medically Actionable Secondary Findings in the 1000 Genomes

The American College of Medical Genetics and Genomics (ACMG) recommends that clinical sequencing laboratories return secondary findings in 56 genes associated with medically actionable conditions. Our goal was to apply a systematic, stringent approach consistent with clinical standards to estimate the prevalence of pathogenic variants associated with such conditions using a diverse sequencing reference sample. Candidate variants in the 56 ACMG genes were selected from Phase 1 of the 1000 Genomes dataset, which contains sequencing information on 1,092 unrelated individuals from across the world. These variants were filtered using the Human Gene Mutation Database (HGMD) Professional version and defined parameters, appraised through literature review, and examined by a clinical laboratory specialist and expert physician. Over 70,000 genetic variants were extracted from the 56 genes, and filtering identified 237 variants annotated as disease causing by HGMD Professional. Literature review and expert evaluation determined that 7 of these variants were pathogenic or likely pathogenic. Furthermore, 5 additional truncating variants not listed as disease causing in HGMD Professional were identified as likely pathogenic. These 12 secondary findings are associated with diseases that could inform medical follow-up, including cancer predisposition syndromes, cardiac conditions, and familial hypercholesterolemia. The majority of the identified medically actionable findings were in individuals from the European (5/379) and Americas (4/181) ancestry groups, with fewer findings in Asian (2/286) and African (1/246) ancestry groups. Our results suggest that medically relevant secondary findings can be identified in approximately 1% (12/1092) of individuals in a diverse reference sample. As clinical sequencing laboratories continue to implement the ACMG recommendations, our results highlight that at least a small number of potentially important secondary findings can be selected for return. Our results also confirm that understudied populations will not reap proportionate benefits of genomic medicine, highlighting the need for continued research efforts on genetic diseases in these populations.     

HLA Typing from 1000 Genomes Whole Genome and Whole Exome Illumina Data

Specific HLA genotypes are known to be linked to either resistance or susceptibility to certain diseases or sensitivity to certain drugs. In addition, high accuracy HLA typing is crucial for organ and bone marrow transplantation. The most widespread high resolution HLA typing method used to date is Sanger sequencing based typing (SBT), and next generation sequencing (NGS) based HLA typing is just starting to be adopted as a higher throughput, lower cost alternative. By HLA typing the HapMap subset of the public 1000 Genomes paired Illumina data, we demonstrate that HLA-A, B and C typing is possible from exome sequencing samples with higher than 90% accuracy. The older 1000 Genomes whole genome sequencing read sets are less reliable and generally unsuitable for the purpose of HLA typing. We also propose using coverage % (the extent of exons covered) as a quality check (QC) measure to increase reliability.     

HLA Diversity in the 1000 Genomes Dataset

The 1000 Genomes Project aims to provide a deep characterization of human genome sequence variation by sequencing at a level that should allow the genome-wide detection of most variants with frequencies as low as 1%. However, in the major histocompatibility complex (MHC), only the top 10 most frequent haplotypes are in the 1% frequency range whereas thousands of haplotypes are present at lower frequencies. Given the limitation of both the coverage and the read length of the sequences generated by the 1000 Genomes Project, the highly variable positions that define HLA alleles may be difficult to identify. We used classical Sanger sequencing techniques to type the HLA-A, HLA-B, HLA-C, HLA-DRB1 and HLA-DQB1 genes in the available 1000 Genomes samples and combined the results with the 103,310 variants in the MHC region genotyped by the 1000 Genomes Project. Using pairwise identity-by-descent distances between individuals and principal component analysis, we established the relationship between ancestry and genetic diversity in the MHC region. As expected, both the MHC variants and the HLA phenotype can identify the major ancestry lineage, informed mainly by the most frequent HLA haplotypes. To some extent, regions of the genome with similar genetic or similar recombination rate have similar properties. An MHC-centric analysis underlines departures between the ancestral background of the MHC and the genome-wide picture. Our analysis of linkage disequilibrium (LD) decay in these samples suggests that overestimation of pairwise LD occurs due to a limited sampling of the MHC diversity. This collection of HLA-specific MHC variants, available on the dbMHC portal, is a valuable resource for future analyses of the role of MHC in population and disease studies.     

Inference of population structure using multilocus genotype data

We describe a model-based clustering method for using multilocus genotype data to infer population structure and assign individuals to populations. We assume a model in which there are K populations (where K may be unknown), each of which is characterized by a set of allele frequencies at each locus. Individuals in the sample are assigned (probabilistically) to populations, or jointly to two or more populations if their genotypes indicate that they are admixed. Our model does not assume a particular mutation process, and it can be applied to most of the commonly used genetic markers, provided that they are not closely linked. Applications of our method include demonstrating the presence of population structure, assigning individuals to populations, studying hybrid zones, and identifying migrants and admixed individuals. We show that the method can produce highly accurate assignments using modest numbers of loci-e.g. , seven microsatellite loci in an example using genotype data from an endangered bird species. The software used for this article is available from http://www.stats.ox.ac.uk/ approximately pritch/home. html.     

Selecting tagging SNPs for association studies using power calculations from genotype data

Recent studies have indicated that linkage disequilibrium (LD) between single nucleotide polymorphism (SNP) markers can be used to derive a reduced set of tagging SNPs (tSNPs) for genetic association studies. Previous strategies for identifying tSNPs have focused on LD measures or haplotype diversity, but the statistical power to detect disease-associated variants using tSNPs in genetic studies has not been fully characterized. We propose a new approach of selecting tSNPs based on determining the set of SNPs with the highest power to detect association. Two-locus genotype frequencies are used in the power calculations. To show utility, we applied this power method to a large number of SNPs that had been genotyped in Caucasian samples. We demonstrate that a significant reduction in genotyping efforts can be achieved although the reduction depends on genotypic relative risk, inheritance mode and the prevalence of disease in the human population. The tSNP sets identified by our method are remarkably robust to changes in the disease model when small relative risk and additive mode of inheritance are employed. We have also evaluated the ability of the method to detect unidentified SNPs. Our findings have important implications in applying tSNPs from different data sources in association studies.     

Family-based association tests using genotype data with uncertainty

Family-based association studies have been widely used to identify association between diseases and genetic markers. It is known that genotyping uncertainty is inherent in both directly genotyped or sequenced DNA variations and imputed data in silico. The uncertainty can lead to genotyping errors and missingness and can negatively impact the power and Type I error rates of family-based association studies even if the uncertainty is independent of disease status. Compared with studies using unrelated subjects, there are very few methods that address the issue of genotyping uncertainty for family-based designs. The limited attempts have mostly been made to correct the bias caused by genotyping errors. Without properly addressing the issue, the conventional testing strategy, i.e. family-based association tests using called genotypes, can yield invalid statistical inferences. Here, we propose a new test to address the challenges in analyzing case-parents data by using calls with high accuracy and modeling genotype-specific call rates. Our simulations show that compared with the conventional strategy and an alternative test, our new test has an improved performance in the presence of substantial uncertainty and has a similar performance when the uncertainty level is low. We also demonstrate the advantages of our new method by applying it to imputed markers from a genome-wide case-parents association study.