Analysis of human triallelic SNPs by next-generation sequencing

A Wright-Fisher graph model and the impact of directional selection on genetic variation

We introduce a multi-allele Wright-Fisher model with mutation and selection such that allele frequencies at a single locus are traced by the path of a hybrid jump-diffusion process. The state space of the process is given by the vertices and edges of a topological graph, i.e. edges are unit intervals. Vertices represent monomorphic population states and positions on the edges mark the biallelic proportions of ancestral and derived alleles during polymorphic segments. In this setting, mutations can only occur at monomorphic loci. We derive the stationary distribution in mutation-selection-drift equilibrium and obtain the expected allele frequency spectrum under large population size scaling. For the extended model with multiple independent loci we derive rigorous upper bounds for a wide class of associated measures of genetic variation. Within this framework we present mathematically precise arguments to conclude that the presence of directional selection reduces the magnitude of genetic variation, as constrained by the bounds for neutral evolution.

Estimates of heterozygosity from single nucleotide polymorphism markers are context-dependent and often wrong

Genetic diversity is frequently described using heterozygosity, particularly in a conservation context. Often, it is estimated using single nucleotide polymorphisms (SNPs); however, it has been shown that heterozygosity values calculated from SNPs can be biased by both study design and filtering parameters. Though solutions have been proposed to address these issues, our own work has found them to be inadequate in some circumstances. Here, we aimed to improve the reliability and comparability of heterozygosity estimates, specifically by investigating how sample size and missing data thresholds influenced the calculation of autosomal heterozygosity (heterozygosity calculated from across the genome, i.e. fixed and variable sites). We also explored how the standard practice of tri- and tetra-allelic site exclusion could bias heterozygosity estimates and influence eventual conclusions relating to genetic diversity. Across three distinct taxa (a frog, Litoria rubella; a tree, Eucalyptus microcarpa; and a grasshopper, Keyacris scurra), we found heterozygosity estimates to be meaningfully affected by sample size and missing data thresholds, partly due to the exclusion of tri- and tetra-allelic sites. These biases were inconsistent both between species and populations, with more diverse populations tending to have their estimates more severely affected, thus having potential to dramatically alter interpretations of genetic diversity. We propose a modified framework for calculating heterozygosity that reduces bias and improves the utility of heterozygosity as a measure of genetic diversity, whilst also highlighting the need for existing population genetic pipelines to be adjusted such that tri- and tetra-allelic sites be included in calculations.

Relatedness coefficients and their applications for triplets and quartets of genetic markers

Relatedness coefficients which seek the identity-by-descent of genetic markers are described. The markers are in groups of two, three or four, and if four, can consist of two pairs. It is essential to use cumulants (not moments) for four-marker-gene probabilities, as the covariance of homozygosity, used in four-marker applications, can only be described with cumulants. A covariance of homozygosity between pairs of markers arises when populations follow a mixture distribution. Also, the probability of four markers all identical-by-descent equals the normalized fourth cumulant. In this article, a "genetic marker" generally represents either a gene locus or an allele at a locus. Applications of three marker coefficients mainly involve conditional regression, and applications of four marker coefficients can involve identity disequilibrium. Estimation of relatedness using genetic marker data is discussed. However, three- and four-marker estimators suffer from statistical and numerical problems, including higher statistical variance, complexity of estimation formula, and singularity at some intermediate allele frequencies.

Genotyping SNPs in lignin biosynthesis gene (CAD1) and transcription factors (MYB1 and MYB2) exhibits association with wood density in teak (Tectona grandis L.f.)

BACKGROUND

Teak (Tectona grandis L.f.), an important source of tropical timber with immense economic value, is a highly outcrossing forest tree species. 150 unrelated accessions of teak (Tectona grandis L.f.) plus trees assembled as clones at National Teak Germplasm Bank, Chandrapur, Maharashtra, India was investigated for association mapping of candidate lignin biosynthesis gene (CAD1) and transcription factors (MYB1 and MYB2).

METHODS AND RESULTS

The CAD1, MYB1 and MYB2 were amplified using specifically designed primers. The amplified sequences were then sequenced and genotyped for 112 SNPs/11 indels. We evaluated the association between SNPs and wood density in teak accessions using GLM and MLM statistical models, with Bonferroni correction applied. The teak accessions recorded an average wood density of 416.69 kg.m-3 (CV 4.97%) and comprised of three loosely structured admixed sub-populations (K = 3), containing 72.05% genetic variation within sub-populations with low intragenic LD (0-21% SNP pairs) at P < 0.05 and high LD decay (33-934 bp) at R2 = 0.1. GLM and MLM models discounting systematic biases (Q and K matrices) to avoid false discovery revealed five loci at rare variants (MAF 0.003) and three loci at common variants (MAF 0.05) to be significantly (P < 0.05) associated with the wood density. However, the stringent Bonferroni correction (4.06-7.04 × 10-4) yielded only a single associated locus (B1485C/A) from exon of MYB1 transcription factor, contributing to about 10.35% phenotypic variation in wood density trait.

CONCLUSION

Scored SNP locus (B1485C/A) can be developed as a molecular probe for selection of improved planting stock with proven wood density trait for a large-scale teak plantation.

Validation of a Large Custom-Designed Pharmacogenomics Panel on an Array Genotyping Platform

BACKGROUND

Pharmacogenomics has the potential to improve patient outcomes through predicting drug response. We designed and evaluated the analytical performance of a custom OpenArray® pharmacogenomics panel targeting 478 single-nucleotide variants (SNVs).

METHODS

Forty Coriell Institute cell line (CCL) DNA samples and DNA isolated from 28 whole-blood samples were used for accuracy evaluation. Genotyping calls were compared to at least 1 reference method: next-generation sequencing, Sequenom MassARRAY®, or Sanger sequencing. For precision evaluation, 23 CCL samples were analyzed 3 times and reproducibility of the assays was assessed. For sensitivity evaluation, 6 CCL samples and 5 whole-blood DNA samples were analyzed at DNA concentrations of 10 ng/µL and 50 ng/µL, and their reproducibility and genotyping call rates were compared.

RESULTS

For 443 variants, all samples assayed had concordant calls with at least 1 reference genotype and also demonstrated reproducibility. However, 6 of these 443 variants showed an unsatisfactory performance, such as low PCR amplification or insufficient separation of genotypes in scatter plots. Call rates were comparable between 50 ng/µL DNA (99.6%) and 10 ng/µL (99.2%). Use of 10 ng/µL DNA resulted in an incorrect call for a single sample for a single variant. Thus, as recommended by the manufacturer, 50 ng/µL is the preferred concentration for patient genotyping.

CONCLUSIONS

We evaluated a custom-designed pharmacogenomics panel and found that it reliably interrogated 437 variants. Clinically actionable results from selected variants on this panel are currently used in clinical studies employing pharmacogenomics for clinical decision-making.

Analysis of human mitochondrial genome co-occurrence networks of Asian population at varying altitudes

Networks have been established as an extremely powerful framework to understand and predict the behavior of many large-scale complex systems. We studied network motifs, the basic structural elements of networks, to describe the possible role of co-occurrence of genomic variations behind high altitude adaptation in the Asian human population. Mitochondrial DNA (mtDNA) variations have been acclaimed as one of the key players in understanding the biological mechanisms behind adaptation to extreme conditions. To explore the cumulative effects of variations in the mitochondrial genome with the variation in the altitude, we investigated human mt-DNA sequences from the NCBI database at different altitudes under the co-occurrence motifs framework. Analysis of the co-occurrence motifs using similarity clustering revealed a clear distinction between lower and higher altitude regions. In addition, the previously known high altitude markers 3394 and 7697 (which are definitive sites of haplogroup M9a1a1c1b) were found to co-occur within their own gene complexes indicating the impact of intra-genic constraint on co-evolution of nucleotides. Furthermore, an ancestral 'RSRS50' variant 10,398 was found to co-occur only at higher altitudes supporting the fact that a separate route of colonization at these altitudes might have taken place. Overall, our analysis revealed the presence of co-occurrence interactions specific to high altitude at a whole mitochondrial genome level. This study, combined with the classical haplogroups analysis is useful in understanding the role of co-occurrence of mitochondrial variations in high altitude adaptation.

HLA-G +3196 polymorphism as a risk factor for cell mediated rejection following heart transplant

BACKGROUND

Rejection is a leading cause of mortality following heart transplantation. Human leukocyte antigen-G (HLA-G) is an immune checkpoint which dampens the immune response. Reports suggest elevated HLA-G expression is associated with reduced allograft rejection. Our objective was to evaluate HLA-G polymorphisms and cell mediated rejection (CMR) development.

METHODS

Recipients (n = 123) were genotyped to identify relevant HLA-G polymorphisms in the 5'regulatory (-725, -201), 3'untranslated (+3197, +3187, +3142, 14-bp indel) and coding regions (haplotypes 1-6). CMR was evaluated via endomyocardial biopsy (grade ≥ 2R). Univariate/adjusted analyses were conducted via Kaplan Meier and proportional hazard models.

RESULTS

Mean recipient age was 48 (±12) years, with a median time to CMR of 4.6 years. 55 (45%) recipients had a biopsy grade ≥ 2R. Adjusted analysis revealed the +3196 G allele as a risk factor for CMR (p = 0.03). Compared to the minor GG genotype, CG had a 47.2% reduction in CMR risk (HR[95% CI] = 0.528 [0.235, 1.184]), while CC had a 66.9% reduction (0.331 [0.144, 0.761]). The recessive effect significantly increased CMR likelihood (2.388 [1.128, 5.059], p = 0.02).

CONCLUSION

The HLA-G +3196 G allele was identified as a risk factor for CMR diagnosis. HLA-G may have a role in therapeutic/diagnostic strategies against transplant rejection.

Empirical design of a variant quality control pipeline for whole genome sequencing data using replicate discordance

The success of next-generation sequencing depends on the accuracy of variant calls. Few objective protocols exist for QC following variant calling from whole genome sequencing (WGS) data. After applying QC filtering based on Genome Analysis Tool Kit (GATK) best practices, we used genotype discordance of eight samples that were sequenced twice each to evaluate the proportion of potentially inaccurate variant calls. We designed a QC pipeline involving hard filters to improve replicate genotype concordance, which indicates improved accuracy of genotype calls. Our pipeline analyzes the efficacy of each filtering step. We initially applied this strategy to well-characterized variants from the ClinVar database, and subsequently to the full WGS dataset. The genome-wide biallelic pipeline removed 82.11% of discordant and 14.89% of concordant genotypes, and improved the concordance rate from 98.53% to 99.69%. The variant-level read depth filter most improved the genome-wide biallelic concordance rate. We also adapted this pipeline for triallelic sites, given the increasing proportion of multiallelic sites as sample sizes increase. For triallelic sites containing only SNVs, the concordance rate improved from 97.68% to 99.80%. Our QC pipeline removes many potentially false positive calls that pass in GATK, and may inform future WGS studies prior to variant effect analysis.

Analysis of genome-wide variants through bulked segregant RNA sequencing reveals a major gene for resistance to Plasmodiophora brassicae in Brassica oleracea

Two cabbage (Brassica oleracea) cultivars 'Tekila' and 'Kilaherb' were identified as resistant to several pathotypes of Plasmodiophora brassicae. In this study, we identified a clubroot resistance gene (Rcr7) in 'Tekila' for resistance to pathotype 3 of P. brassicae from a segregating population derived from 'Tekila' crossed with the susceptible line T010000DH3. Genetic mapping was performed by identifying the percentage of polymorphic variants (PPV), a new method proposed in this study, through bulked segregant RNA sequencing. Chromosome C7 carried the highest PPV (42%) compared to the 30-34% in the remaining chromosomes. A peak with PPV (56-73%) was found within the physical interval 41-44 Mb, which indicated that Rcr7 might be located in this region. Kompetitive Allele-Specific PCR was used to confirm the association of Rcr7 with SNPs in the region. Rcr7 was flanked by two SNP markers and co-segregated with three SNP markers in the segregating population of 465 plants. Seven genes encoding TIR-NBS-LRR disease resistance proteins were identified in the target region, but only two genes, Bo7g108760 and Bo7g109000, were expressed. Resistance to pathotype 5X was also mapped to the same region as Rcr7. B. oleracea lines including 'Kilaherb' were tested with five SNP markers for Rcr7 and for resistance to pathotype 3; 11 of 25 lines were resistant, but 'Kilaherb' was the only line that carried the SNP alleles associated with Rcr7. The presence of Rcr7 in 'Kilaherb' for resistance to both pathotypes 3 and 5X was confirmed through linkage analysis.

Genotype Calling from Population-Genomic Sequencing Data

Genotype calling plays important roles in population-genomic studies, which have been greatly accelerated by sequencing technologies. To take full advantage of the resultant information, we have developed maximum-likelihood (ML) methods for calling genotypes from high-throughput sequencing data. As the statistical uncertainties associated with sequencing data depend on depths of coverage, we have developed two types of genotype callers. One approach is appropriate for low-coverage sequencing data, and incorporates population-level information on genotype frequencies and error rates pre-estimated by an ML method. Performance evaluation using computer simulations and human data shows that the proposed framework yields less biased estimates of allele frequencies and more accurate genotype calls than current widely used methods. Another type of genotype caller applies to high-coverage sequencing data, requires no prior genotype-frequency estimates, and makes no assumption on the number of alleles at a polymorphic site. Using computer simulations, we determine the depth of coverage necessary to accurately characterize polymorphisms using this second method. We applied the proposed method to high-coverage (mean 18×) sequencing data of 83 clones from a population of Daphnia pulex. The results show that the proposed method enables conservative and reasonably powerful detection of polymorphisms with arbitrary numbers of alleles. We have extended the proposed method to the analysis of genomic data for polyploid organisms, showing that calling accurate polyploid genotypes requires much higher coverage than diploid genotypes.

Rare Loss-of-Function Variants in NPC1 Predispose to Human Obesity

Some Shanghai Clinical Center f a role of Niemann-Pick type C1 (NPC1) for obesity traits. However, whether the loss-of-function mutations in NPC1 cause adiposity in humans remains unknown. We recruited 25 probands with rare autosomal-recessive Niemann-Pick type C (NP-C) disease and their parents in assessment of the effect of heterozygous NPC1 mutations on adiposity. We found that male NPC1^+/- carriers had a significantly higher BMI than matched control subjects or the whole population-based control subjects. Consistently, male NPC1^+/- mice had increased fat storage while eating a high-fat diet. We further conducted an in-depth assessment of rare variants in the NPC1 gene in young, severely obese subjects and lean control subjects and identified 17 rare nonsynonymous/frameshift variants in NPC1 (minor allele frequency <1%) that were significantly associated with an increased risk of obesity (3.40% vs. 0.73%, respectively, in obese patients and control subjects, P = 0.0008, odds ratio = 4.8, 95% CI 1.7-13.2), indicating that rare NPC1 variants were enriched in young, morbidly obese Chinese subjects. Importantly, participants carrying rare variants with severely damaged cholesterol-transporting ability had more fat accumulation than those with mild/no damage rare variants. In summary, rare loss-of-function NPC1 mutations were identified as being associated with human adiposity with a high penetrance, providing potential therapeutic interventions for obesity in addition to the role of NPC1 in the familial NP-C disease.