A rank-based nonparametric method for mapping quantitative trait loci in outbred half-sib pedigrees: application to milk production in a granddaughter design

Comparing linkage and association analyses in sheep points to a better way of doing GWAS

Genome wide association studies (GWAS) have largely succeeded family-based linkage studies in livestock and human populations as the preferred method to map loci for complex or quantitative traits. However, the type of results produced by the two analyses contrast sharply due to differences in linkage disequilibrium (LD) imposed by the design of studies. In this paper, we demonstrate that association and linkage studies are in agreement provided that (i) the effects from both studies are estimated appropriately as random effects, (ii) all markers are fitted simultaneously and (iii) appropriate adjustments are made for the differences in LD between the study designs. We demonstrate with real data that linkage results can be predicted by the sum of association effects. Our association study captured most of the linkage information because we could predict the linkage results with moderate accuracy. We suggest that the ability of common single nucleotide polymorphism (SNP) to capture the genetic variance in a population will depend on the effective population size of the study organism. The results provide further evidence for many loci of small effect underlying complex traits. The analysis suggests a more informed method for GWAS is to fit statistical models where all SNPs are analysed simultaneously and as random effects.

Rank regression for analyzing ordinal qualitative data for treatment comparison

ABSTRACT Ordinal qualitative data are often collected for phenotypical measurements in plant pathology and other biological sciences. Statistical methods, such as t tests or analysis of variance, are usually used to analyze ordinal data when comparing two groups or multiple groups. However, the underlying assumptions such as normality and homogeneous variances are often violated for qualitative data. To this end, we investigated an alternative methodology, rank regression, for analyzing the ordinal data. The rank-based methods are essentially based on pairwise comparisons and, therefore, can deal with qualitative data naturally. They require neither normality assumption nor data transformation. Apart from robustness against outliers and high efficiency, the rank regression can also incorporate covariate effects in the same way as the ordinary regression. By reanalyzing a data set from a wheat Fusarium crown rot study, we illustrated the use of the rank regression methodology and demonstrated that the rank regression models appear to be more appropriate and sensible for analyzing nonnormal data and data with outliers.

Variants modulating the expression of a chromosome domain encompassing PLAG1 influence bovine stature

We report mapping of a quantitative trait locus (QTL) with a major effect on bovine stature to a ∼780-kb interval using a Hidden Markov Model-based approach that simultaneously exploits linkage and linkage disequilibrium. We re-sequenced the interval in six sires with known QTL genotype and identified 13 clustered candidate quantitative trait nucleotides (QTNs) out of >9,572 discovered variants. We eliminated five candidate QTNs by studying the phenotypic effect of a recombinant haplotype identified in a breed diversity panel. We show that the QTL influences fetal expression of seven of the nine genes mapping to the ∼780-kb interval. We further show that two of the eight candidate QTNs, mapping to the PLAG1-CHCHD7 intergenic region, influence bidirectional promoter strength and affect binding of nuclear factors. By performing expression QTL analyses, we identified a splice site variant in CHCHD7 and exploited this naturally occurring null allele to exclude CHCHD7 as single causative gene.

Mapping genome-wide QTL of ratio traits with Bayesian shrinkage analysis for its component traits

The ratio trait is defined as a ratio of two regular quantitative traits with normal distribution, which is distinguished from regular quantitative traits in the genetic analysis because it does not follow the normal distribution. On the basis of maximum likelihood method that uses a special linear combination of the two component traits, we develop a Bayesian mapping strategy for ratio traits, which firstly analyzes the two component traits by Bayesian shrinkage method, and then generates a new posterior sample of genetic effects for a ratio trait from ones of population means and genetic effects for the two component traits, finally, infers QTL for the ratio trait via post MCMC analysis for the new posterior sample. A simulation study demonstrates that the new method has higher detecting power of the QTL than maximum likelihood method. An application is illustrated to map genome-wide QTL for relative growth rate of height on soybean.

Quantitative trait loci for resistance to trichostrongylid infection in Spanish Churra sheep

BACKGROUND

For ruminants reared on grazing systems, gastrointestinal nematode (GIN) parasite infections represent the class of diseases with the greatest impact on animal health and productivity. Among the many possible strategies for controlling GIN infection, the enhancement of host resistance through the selection of resistant animals has been suggested by many authors. Because of the difficulty of routinely collecting phenotypic indicators of parasite resistance, information derived from molecular markers may be used to improve the efficiency of classical genetic breeding.

METHODS

A total of 181 microsatellite markers evenly distributed along the 26 sheep autosomes were used in a genome scan analysis performed in a commercial population of Spanish Churra sheep to detect chromosomal regions associated with parasite resistance. Following a daughter design, we analysed 322 ewes distributed in eight half-sib families. The phenotypes studied included two faecal egg counts (LFEC0 and LFEC1), anti-Teladorsagia circumcincta LIV IgA levels (IgA) and serum pepsinogen levels (Peps).

RESULTS

The regression analysis revealed one QTL at the 5% genome-wise significance level on chromosome 6 for LFEC1 within the marker interval BM4621-CSN3. This QTL was found to be segregating in three out of the eight families analysed. Four other QTL were identified at the 5% chromosome-wise level on chromosomes 1, 10 and 14. Three of these QTL influenced faecal egg count, and the other one had an effect on IgA levels.

CONCLUSION

This study has successfully identified segregating QTL for parasite resistance traits in a commercial population. For some of the QTL detected, we have identified interesting coincidences with QTL previously reported in sheep, although most of those studies have been focused on young animals. Some of these coincidences might indicate that some common underlying loci affect parasite resistance traits in different sheep breeds. The identification of new QTL may suggest the existence of complex host-parasite relationships that have unique features depending on the host-parasite combination, perhaps due to the different mechanisms underlying resistance in adult sheep (hypersensitivity reactions) and lambs (immunity). The most significant QTL identified on chromosome 6 for LFEC(1) may be the target for future fine-mapping research efforts.

Mapping QTL influencing gastrointestinal nematode burden in Dutch Holstein-Friesian dairy cattle

Background

Parasitic gastroenteritis caused by nematodes is only second to mastitis in terms of health costs to dairy farmers in developed countries. Sustainable control strategies complementing anthelmintics are desired, including selective breeding for enhanced resistance.

Results and Conclusion

To quantify and characterize the genetic contribution to variation in resistance to gastro-intestinal parasites, we measured the heritability of faecal egg and larval counts in the Dutch Holstein-Friesian dairy cattle population. The heritability of faecal egg counts ranged from 7 to 21% and was generally higher than for larval counts. We performed a whole genome scan in 12 paternal half-daughter groups for a total of 768 cows, corresponding to the ~10% most and least infected daughters within each family (selective genotyping). Two genome-wide significant QTL were identified in an across-family analysis, respectively on chromosomes 9 and 19, coinciding with previous findings in orthologous chromosomal regions in sheep. We identified six more suggestive QTL by within-family analysis. An additional 73 informative SNPs were genotyped on chromosome 19 and the ensuing high density map used in a variance component approach to simultaneously exploit linkage and linkage disequilibrium in an initial inconclusive attempt to refine the QTL map position.

Robust Bayesian mapping of quantitative trait loci using Student-t distribution for residual

In most quantitative trait loci (QTL) mapping studies, phenotypes are assumed to follow normal distributions. Deviations from this assumption may affect the accuracy of QTL detection, leading to detection of false positive QTL. To improve the robustness of QTL mapping methods, we replace the normal distribution assumption for residuals in a multiple QTL model with a Student-t distribution that is able to accommodate residual outliers. A Robust Bayesian mapping strategy is proposed on the basis of the Bayesian shrinkage analysis for QTL effects. The simulations show that Robust Bayesian mapping approach can substantially increase the power of QTL detection when the normality assumption does not hold and applying it to data already normally distributed does not influence the result. The proposed QTL mapping method is applied to mapping QTL for the traits associated with physics-chemical characters and quality in rice. Similarly to the simulation study in the real data case the robust approach was able to detect additional QTLs when compared to the traditional approach. The program to implement the method is available on request from the first or the corresponding author.

Bayesian robust analysis for genetic architecture of quantitative traits

Motivation: In most quantitative trait locus (QTL) mapping studies, phenotypes are assumed to follow normal distributions. Deviations from this assumption may affect the accuracy of QTL detection and lead to detection of spurious QTLs. To improve the robustness of QTL mapping methods, we replaced the normal distribution for residuals in multiple interacting QTL models with the normal/independent distributions that are a class of symmetric and long-tailed distributions and are able to accommodate residual outliers. Subsequently, we developed a Bayesian robust analysis strategy for dissecting genetic architecture of quantitative traits and for mapping genome-wide interacting QTLs in line crosses.

Results: Through computer simulations, we showed that our strategy had a similar power for QTL detection compared with traditional methods assuming normal-distributed traits, but had a substantially increased power for non-normal phenotypes. When this strategy was applied to a group of traits associated with physical/chemical characteristics and quality in rice, more main and epistatic QTLs were detected than traditional Bayesian model analyses under the normal assumption.

Contact:runqingyang@sjtu.edu.cn; dengh@umkc.edu

Supplementary information:Supplementary data are available at Bioinformatics online.

Chromosomal regions underlying noncoagulation of milk in Finnish Ayrshire cows

About 10% of Finnish Ayrshire cows produce noncoagulating milk, i.e., milk that does not form a curd in a standard 30-min testing time and is thus a poor raw material for cheese dairies. This phenomenon is associated with peak and midlactation, but some cows produce noncoagulating milk persistently. A genomewide scan under a selective DNA pooling method was carried out to locate genomic regions associated with the noncoagulation of milk. On the basis of the hypothesis of the same historical mutation, we pooled the data across sires. Before testing pools for homogeneity, allele intensities were corrected for PCR artifacts, i.e., shadow bands and differential amplification. Results indicating association were verified using daughter design and selective genotyping within families. Data consisted of 18 sire families with 477 genotyped daughters in total, i.e., 12% of each tail of the milk coagulation ability. Data were analyzed using interval mapping under maximum-likelihood and nonparametric methods. BMS1126 on chromosome 2 and BMS1355 on chromosome 18 were associated with noncoagulation of milk across families on an experimentwise 0.1% significance level. By scanning gene databases, we found two potential candidate genes: LOC538897, a nonspecific serine/threonine kinase on chromosome 2, and SIAT4B, a sialyltransferase catalyzing the last step of glycosylation of kappa-casein on chromosome 18. Further studies to determine the role of the candidates in the noncoagulation of milk are clearly needed.

Short communication: detection of quantitative trait loci influencing somatic cell score in Spanish Churra sheep

Eleven half-sib ovine families, including 1,421 Spanish Churra ewes, were analyzed for 181 microsatellite markers spanning the entire autosomic ovine genome. Using a multimarker regression method, a daughter experimental design was used to identify putative quantitative trait loci (QTL) affecting the somatic cell score (SCS). Chromosome-wise significance thresholds were set empirically by permuting the phenotypic data. Marker order and genetic distances of the autosomic linkage map built for this commercial population were in accordance with the published ovine linkage map. An across-family association analysis revealed a region on chromosome 20 suggestive of evidence for a QTL. Segregation of the QTL into 2 families was inferred from the within-family analysis, and differences in the position of the suggested QTL were found between the 2 half-sib groups. This could be the result of incomplete information associated with the markers for the significant families. The location of the major histocompatibility complex in proximity to the across-family effect suggests this region may harbor a segregating QTL for the SCS in the Churra population. Studies in dairy cattle examining the SCS have reported linkage associations on corresponding bovine orthologous regions, supporting the validity of our findings.

Nonparametric functional mapping of quantitative trait loci

Functional mapping is a useful tool for mapping quantitative trait loci (QTL) that control dynamic traits. It incorporates mathematical aspects of biological processes into the mixture model-based likelihood setting for QTL mapping, thus increasing the power of QTL detection and the precision of parameter estimation. However, in many situations there is no obvious functional form and, in such cases, this strategy will not be optimal. Here we propose to use nonparametric function estimation, typically implemented with B-splines, to estimate the underlying functional form of phenotypic trajectories, and then construct a nonparametric test to find evidence of existing QTL. Using the representation of a nonparametric regression as a mixed model, the final test statistic is a likelihood ratio test. We consider two types of genetic maps: dense maps and general maps, and the power of nonparametric functional mapping is investigated through simulation studies and demonstrated by examples.

Box-Cox transformation for QTL mapping

The maximum likelihood method of QTL mapping assumes that the phenotypic values of a quantitative trait follow a normal distribution. If the assumption is violated, some forms of transformation should be taken to make the assumption approximately true. The Box-Cox transformation is a general transformation method which can be applied to many different types of data. The flexibility of the Box-Cox transformation is due to a variable, called transformation factor, appearing in the Box-Cox formula. We developed a maximum likelihood method that treats the transformation factor as an unknown parameter, which is estimated from the data simultaneously along with the QTL parameters. The method makes an objective choice of data transformation and thus can be applied to QTL analysis for many different types of data. Simulation studies show that (1) Box-Cox transformation can substantially increase the power of QTL detection; (2) Box-Cox transformation can replace some specialized transformation methods that are commonly used in QTL mapping; and (3) applying the Box-Cox transformation to data already normally distributed does not harm the result.

Results of a whole-genome quantitative trait locus scan for growth, carcass composition and meat quality in a porcine four-way cross

A whole-genome quantitative trait locus (QTL) scan for 31 phenotypes related to growth, carcass composition and meat quality was conducted using 1187 progeny of a commercial four-way cross. Animals were genotyped for 198 microsatellite markers that spanned the entire porcine genome. QTL analysis was conducted to extract information from paternal and maternal meioses separately using a rank-based nonparametric approach for half-sib designs. Nine QTL exceeded genome-wide significance: one QTL affecting growth (average daily gain on SSC1), two QTL influencing carcass composition (fatness on SSC3 and muscle mass on SSC15) and six QTL influencing meat quality (tenderness on SSC4 and SSC14; colour on SSC5, SSC6 and SSCX; and conductivity on SSC16). All but one of these coincided with previously reported QTL. In addition, we present evidence for 78 suggestive QTL with a combined false discovery rate of 40%.

Mapping of quantitative trait loci for carcass traits in a Japanese Black (Wagyu) cattle population

To detect quantitative trait loci (QTL) that influence economically important traits in a purebred Japanese Black cattle population, we performed a preliminary genome-wide scan using 187 microsatellite markers across a paternal half-sib family composed of 258 offspring. We located six QTL at the 1% chromosome-wise level on bovine chromosomes (BTA) 4, 6, 13, 14 and 21. A second screen of these six QTL regions using 138 additional paternal offspring half-sib from the same sire, provided further support for five QTL: carcass weight on BTA14 (22-39 cM), one for rib thickness on BTA6 (27-58 cM) and three for beef marbling score (BMS) on BTA4 (59-67 cM), BTA6 (68-89 cM) and BTA21 (75-84 cM). The location of QTL for subcutaneous fat thickness on BTA13 was not supported by the second screen (P > 0.05). We determined that the combined contribution of the three QTLs for BMS was 10.1% of the total variance. The combined phenotypic average of these three Q was significantly different (P < 0.001) from those of other allele combinations. Analysis of additional half-sib families will be necessary to confirm these QTL.

Genetical metabolomics of flavonoid biosynthesis in Populus: a case study

Genetical metabolomics [metabolite profiling combined with quantitative trait locus (QTL) analysis] has been proposed as a new tool to identify loci that control metabolite abundances. This concept was evaluated in a case study with the model tree Populus. Using HPLC, the peak abundances were analyzed of 15 closely related flavonoids present in apical tissues of two full-sib poplar families, Populus deltoides cv. S9-2 x P. nigra cv. Ghoy and P. deltoides cv. S9-2 x P. trichocarpa cv. V24, and correlation and QTL analysis were used to detect flux control points in flavonoid biosynthesis. Four robust metabolite quantitative trait loci (mQTL), associated with rate-limiting steps in flavonoid biosynthesis, were mapped. Each mQTL was involved in the flux control to one or two flavonoids. Based on the identities of the affected metabolites and the flavonoid pathway structure, a tentative function was assigned to three of these mQTL, and the corresponding candidate genes were mapped. The data indicate that the combination of metabolite profiling with QTL analysis is a valuable tool to identify control points in a complex metabolic pathway of closely related compounds.

Influence of Prion Protein Genotypes on Milk Production Traits in Spanish Churra Sheep

The aim of this work was to analyze the possible relationships between milk production traits and prion protein genotypes in Spanish Churra sheep. For this purpose, 2 analyses were carried out. First, an association study was performed of the prion protein genotypes of 12,533 Churra ewes and their milk yield, protein percentage, fat percentage, and somatic cell score as phenotypes, followed by a quantitative trait loci screening on the chromosome where the prion protein gene was located in this population. The latter analysis was carried out using 8 genetic markers (7 microsatellites and the prion protein genotypes) spanning ovine chromosome 13 using a daughter design. Regarding genotype frequencies, the most frequent allele was ARQ (75.90%), which linked with a high susceptibility to scrapie, followed by the resistant haplotype, ARR (18.16 %). The frequency of the most susceptible allele, VRQ, was around 1%. No evidence of association or linkage between prion protein genotypes and milk traits has been detected in Churra sheep. These results indicate that increasing the ARR frequency in Churra population will not have an adverse effect on selection for milk traits included in the breeding objectives. However, the low allele frequencies for ARR should be considered in the initial stages to prevent possible bottlenecks in future genetic progress.

Genetic identification of distinct loci controlling mammary tumor multiplicity, latency, and aggressiveness in the rat

The rat is considered an excellent model for studying human breast cancer. Therefore, understanding the genetic basis of susceptibility to mammary cancer in this species is of great interest. Previous studies based on crosses involving the susceptible strain WF (crossed with the resistant strains COP or WKY) and focusing on tumor multiplicity as the susceptibility phenotype led to the identification of several loci that control chemically induced mammary cancer. The present study was aimed to determine whether other loci can be identified by analyzing crosses derived from another susceptible strain on the one hand, and by including phenotypes other than tumor multiplicity on the other hand. A backcross was generated between the susceptible SPRD-Cu3 strain and the resistant WKY strain. Female progeny were genotyped with microsatellite markers covering all rat autosomes, treated with a single dose of DMBA, and phenotyped with respect to tumor latency, tumor multiplicity, and tumor aggressiveness. Seven loci controlling mammary tumor development were detected. Different loci control tumor multiplicity, latency, and aggressiveness. While some of these loci colocalize with loci identified in crosses involving the susceptible strain WF, new loci have been uncovered, indicating that the use of distinct susceptible and resistant strain pairs will help in establishing a comprehensive inventory of mammary cancer susceptibility loci.

Molecular dissection of a quantitative trait locus: a phenylalanine-to-tyrosine substitution in the transmembrane domain of the bovine growth hormone receptor is associated with a major effect on milk yield and composition

We herein report on our efforts to improve the mapping resolution of a QTL with major effect on milk yield and composition that was previously mapped to bovine chromosome 20. By using a denser chromosome 20 marker map and by exploiting linkage disequilibrium using two distinct approaches, we provide strong evidence that a chromosome segment including the gene coding for the growth hormone receptor accounts for at least part of the chromosome 20 QTL effect. By sequencing individuals with known QTL genotype, we identify an F to Y substitution in the transmembrane domain of the growth hormone receptor gene that is associated with a strong effect on milk yield and composition in the general population.

Positional cloning of the gene LIMBIN responsible for bovine chondrodysplastic dwarfism

Chondrodysplastic dwarfism in Japanese brown cattle is an autosomal recessive disorder characterized by short limbs. Previously, we mapped the locus responsible for the disease on the distal end of bovine chromosome 6. Here, we narrowed the critical region to approximately 2 cM by using linkage analysis, constructed a BAC and YAC contig covering this region, and identified a gene, LIMBIN (LBN), that possessed disease-specific mutations in the affected calves. One mutation was a single nucleotide substitution leading to an activation of a cryptic splicing donor site and the other was a one-base deletion resulting in a frameshift mutation. Strong expression of the Lbn gene was observed in limb buds of developing mouse embryos and in proliferating chondrocytes and bone-forming osteoblasts in long bones. These findings indicate that LBN is responsible for bovine chondrodysplastic dwarfism and has a critical role in a skeletal development.

Simultaneous mining of linkage and linkage disequilibrium to fine map quantitative trait loci in outbred half-sib pedigrees: revisiting the location of a quantitative trait locus with major effect on milk production on bovine chromosome 14

A maximum-likelihood QTL mapping method that simultaneously exploits linkage and linkage disequilibrium and that is applicable in outbred half-sib pedigrees is described. The method is applied to fine map a QTL with major effect on milk fat content in a 3-cM marker interval on proximal BTA14. This proximal location is confirmed by applying a haplotype-based association method referred to as recombinant ancestral haplotype analysis. The origin of the discrepancy between the QTL position derived in this work and that of a previous analysis is examined and shown to be due to the existence of distinct marker haplotypes associated with QTL alleles having large substitution effects.

Robust QTL effect estimation using the minimum distance method

Robustness has received little attention in QTL studies. We compare Maximum Likelihood (ML) and the Minimum Distance (MD) methods when there exists data contamination caused by outliers. A backcross population of size (N) 200 and 500 and 0, 5 or 25 outliers was simulated. The mean and standard deviation of the first QTL genotype were set to 1. Four cases were considered: (i) micro2=1, sigma2=1; (ii)micro2=1, sigma2=1.25; (iii) micro2=1.252, sigma2=1; (iv) micro2=1.282, sigma2=1.25, where micro2 and sigma2 are the mean and standard deviation of the second genotype. Either full or selective genotyping was considered. A Monte Carlo MD method is proposed to deal with missing genotypes. MD estimates were much more robust than ML estimates, especially with respect to scale parameter estimates, and with selective genotyping.

Fine-mapping of quantitative trait loci by identity by descent in outbred populations: application to milk production in dairy cattle

We previously mapped a quantitative trait locus (QTL) affecting milk production to bovine chromosome 14. To refine the map position of this QTL, we have increased the density of the genetic map of BTA14q11-16 by addition of nine microsatellites and three single nucleotide polymorphisms. Fine-mapping of the QTL was accomplished by a two-tiered approach. In the first phase, we identified seven sires heterozygous "Qq" for the QTL by marker-assisted segregation analysis in a Holstein-Friesian pedigree comprising 1,158 individuals. In a second phase, we genotyped the seven selected sires for the newly developed high-density marker map and searched for a shared haplotype flanking an hypothetical, identical-by-descent QTL allele with large substitution effect. The seven chromosomes increasing milk fat percentage were indeed shown to carry a common chromosome segment with an estimated size of 5 cM predicted to contain the studied QTL. The same haplotype was shown to be associated with increased fat percentage in the general population as well, providing additional support in favor of the location of the QTL within the corresponding interval.