Estimation of pedigree errors in the UK dairy population using microsatellite markers and the impact on selection

Short communication: evaluation of bovine milk residues from routine milk testing programs as DNA source for genotyping

Genome-wide association studies and genomic evaluation using a dense set of genetic markers both require a large number of genotyped individuals. Collection of the respective samples contributes substantially to the cost of the approach. In dairy cattle research, the use of residues from routine milk recording would be a cost-saving alternative to obtain samples for an appropriate number of individuals with specific phenotypes in a very short time. To assess the suitability of milk recording residues, we concurrently investigated milk residues obtained after standardized milk recording procedures and blood samples from 115 cows originating from 3 farms with different milking systems by genotyping 15 microsatellite markers. We found that 4% of the milk samples were possibly assigned to the wrong animal (i.e., conflicts) and that at least 27% of the milk residues were contaminated, as indicated by an extra allele not present in the blood sample. These additional alleles primarily originated from a sample with a higher somatic cell score that went through the milk sample analyzer in the milk laboratory before the target sample. Furthermore, additional allele carryover was observed across more than one sample, when the difference in somatic cell count between samples exceeded 100,000 cells/mL. Finally, in several samples, the extra allele could not be traced back to previous samples passing through the milk sample analyzer. One source of those contaminations might be sample collection on-farm due to milk traces from the previously milked cow in the hose. No correlation was found between the farm management and conflicts or contaminations. We conclude that residues from routine milk recording are not suitable for genomic evaluation or genome-wide association studies because of the high prevalence of contamination generated at several steps during the collection and processing of milk residual samples.

Power of exclusion of 19 microsatellite markers for parentage testing in river buffalo (Bubalus bubalis)

In the present study, 19 microsatellite markers were assessed for their power of exclusion to test parentage in river buffalo. Microsatellite genotypes of 216 unrelated buffaloes belonging to five different breeds were utilized for the study. The probabilities of exclusion were calculated for three hypothetical situations viz. paternity testing (PE1), one parental genotype unavailable (PE2) and exclusion of both parents i.e. substituted offspring (PE3). The mean probability of exclusion across 19 investigated markers in buffalo was 0.578 (PE1), 0.405 (PE2) and 0.764 (PE3) respectively. The probability of exclusion for paternity (PE1) ranged between 0.297 and 0.814 across different markers. The exclusion probability for the cases one parent unavailable (PE2) and substituted offspring (PE3) varied from 0.143 to 0.688 and 0.465 to 0.946 respectively. Polymorphism information content and expected heterozygosity were found to have significantly high correlation with probability of exclusion of microsatellite markers. The cumulative PE1 of nine marker loci was estimated to be 0.9999 while in case of absence of one of the parental genotypes, a minimum of 11 markers were required to achieve a cumulative PE2 of 0.999. In conclusion, the present study proposes two multiplex sets with four and five markers respectively for routine parentage testing in buffalo and an additional set of four markers for doubtful cases of paternity.

Identification of Mendelian inconsistencies between SNP and pedigree information of sibs

BACKGROUND

Using SNP genotypes to apply genomic selection in breeding programs is becoming common practice. Tools to edit and check the quality of genotype data are required. Checking for Mendelian inconsistencies makes it possible to identify animals for which pedigree information and genotype information are not in agreement.

METHODS

Straightforward tests to detect Mendelian inconsistencies exist that count the number of opposing homozygous marker (e.g. SNP) genotypes between parent and offspring (PAR-OFF). Here, we develop two tests to identify Mendelian inconsistencies between sibs. The first test counts SNP with opposing homozygous genotypes between sib pairs (SIBCOUNT). The second test compares pedigree and SNP-based relationships (SIBREL). All tests iteratively remove animals based on decreasing numbers of inconsistent parents and offspring or sibs. The PAR-OFF test, followed by either SIB test, was applied to a dataset comprising 2,078 genotyped cows and 211 genotyped sires. Theoretical expectations for distributions of test statistics of all three tests were calculated and compared to empirically derived values. Type I and II error rates were calculated after applying the tests to the edited data, while Mendelian inconsistencies were introduced by permuting pedigree against genotype data for various proportions of animals.

RESULTS

Both SIB tests identified animal pairs for which pedigree and genomic relationships could be considered as inconsistent by visual inspection of a scatter plot of pairwise pedigree and SNP-based relationships. After removal of 235 animals with the PAR-OFF test, SIBCOUNT (SIBREL) identified 18 (22) additional inconsistent animals.Seventeen animals were identified by both methods. The numbers of incorrectly deleted animals (Type I error), were equally low for both methods, while the numbers of incorrectly non-deleted animals (Type II error), were considerably higher for SIBREL compared to SIBCOUNT.

CONCLUSIONS

Tests to remove Mendelian inconsistencies between sibs should be preceded by a test for parent-offspring inconsistencies. This parent-offspring test should not only consider parent-offspring pairs based on pedigree data, but also those based on SNP information. Both SIB tests could identify pairs of sibs with Mendelian inconsistencies. Based on type I and II error rates, counting opposing homozygotes between sibs (SIBCOUNT) appears slightly more precise than comparing genomic and pedigree relationships (SIBREL) to detect Mendelian inconsistencies between sibs.

A single nucleotide polymorphism set for paternal identification to reduce the costs of trait recording in commercial pig breeding

In animal breeding, recording of correct pedigrees is essential to achieve genetic progress. Markers on DNA are useful to verify the on-farm pedigree records (parental verification) but can also be used to assign parents retrospectively (parental identification). This approach could reduce the costs of recording for traits with low incidence, such as those related to diseases or mortality. In this study, SNP were used to assign the true sires of 368 purebred animals from a Duroc-based sire line and 140 crossbred offspring from a commercial pig population. Some of the sires were closely related. There were 3 full sibs and 17 half sibs among the true fathers and 4 full sibs and 35 half sibs among all putative fathers. To define the number of SNP necessary, 5 SNP panels (40, 60, 80, 100, and 120 SNP) were assembled from the Illumina PorcineSNP60 Beadchip (Illumina, San Diego, CA) based on minor allele frequency (>0.3), high genotyping call rate (≥90%), and equal spacing across the genome. For paternal identification considering only the 66 true sires in the data set, 60 SNP resulted in 100% correct assignment of the sire. By including additional putative sires (n = 304), 80 SNP were sufficient for 100% correct assignment of the sire. The following criteria were derived to identify the correct sire for the current data set: the logarithm of odds (LOD) score for assigning the correct sire was ≥5, the number of mismatches was ≤1, and the difference in the LOD score between the first and the second most likely sire was >5. If the correct sire was not present among all putative sires, the mean LOD for the most likely sire was close to zero or negative when using 100 SNP. More SNP would be needed for paternal identification if the number of putative sires increased and the degree of relatedness was greater than in the data set used here. The threshold for the number of mismatches can be adjusted according to the practical situation to account for the trade-off between false negatives and false positives. The latter can be avoided efficiently, ensuring that the correct father is being sampled. Nevertheless, a restriction on the number of putative sires is advisable to reduce the risk of assigning close relatives.

Impact of depth of pedigree and inclusion of historical data on the estimation of additive variance and breeding values in a sugarcane breeding program

Sugarcane breeders in Australia combine data across four selection programs to obtain estimates of breeding value for parents. When these data are combined with full pedigree information back to founding parents, computing limitations mean it is not possible to obtain information on all parents. Family data from one sugarcane selection program were analysed using two different genetic models to investigate how different depths of pedigree and amount of data affect the reliability of estimating breeding value of sugarcane parents. These were the parental and animal models. Additive variance components and breeding values estimated from different amounts of information were compared for both models. The accuracy of estimating additive variance components and breeding values improved as more pedigree information and historical data were included in analyses. However, adding years of data had a much larger effect on the estimation of variance components of the population, and breeding values of the parents. To accurately estimate breeding values of all sugarcane parents, a minimum of three generations of pedigree and 5 years of historical data were required, while more information (four generations of pedigree and 7 years of historical data) was required when identifying top parents to be selected for future cross pollination.

Establishment of paternity testing system using microsatellite markers in Chinese Holstein

To estimate the efficiency of microsatellite markers in paternity testing among Chinese Holstein, 30 microsatellite loci were used to differentiate 330 Chinese Holstein genotypes, according to the calculation of the allele frequency, number of alleles, effective number of alleles, genetic heterozygosity, polymorphic information content (PIC), and the exclusion probability in this cattle population. The results demonstrated that the exclusion probability ranged from 0.620 in locus BM1818 to 0.265 in locus INRA005 with the average of 0.472 and 11 microsatellite markers exceeding 0.5. The combined exclusion probability of nine microsatellite markers was over 0.99. The result showed that paternity testing of Chinese Holstein was basically resolved using the nine microsatellite markers selected.

Correlation analysis of economic traits in Liaoning new breed of cashmere goats using microsatellite DNA markers

The economic traits of Liaoning new breed of cashmere goats, a special Chinese genetic resource, were analyzed in 150 animals by typing 11 microsatellite loci. The association between three economic traits (body weight, cashmere yield and fineness) in this new breed of goats and the marker genotypes were analyzed. The results show that: AA and BC at LSCV13, DE at IDVGA64 and BB at BMS2782 were favorable genotypes for body weight. AB at LSCV13, AD and BE at CSSM11, BD and CC at IDVGA64, BC and DE at BMS2782 were favorable genotypes for cashmere yield. AA at CSSM11, BC and DE at IDVGA64, CD at BMS2782 were favorable genotypes for cashmere fineness.

Implementation of a parentage control system in Portuguese beef-cattle with a panel of microsatellite markers

A study was conducted to assess the feasibility of applying a panel of 10 microsatellite markers in parentage control of beef cattle in Portugal. In the first stage, DNA samples were collected from 475 randomly selected animals of the Charolais, Limousin and Preta breeds. Across breeds and genetic markers, means for average number of alleles, effective number of alleles, expected heterozygosity and polymorphic information content, were 8.20, 4.43, 0.733 and 0.70, respectively. Enlightenment from the various markers differed among breeds, but the set of 10 markers resulted in a combined probability above 0.9995 in the ability to exclude a random putative parent. The marker-set thus developed was later used for parentage control in a group of 140 calves from several breeds, where there was the suspicion of possible faulty parentage recording. Overall, 76.4% of the calves in this group were compatible with the recorded parents, with most incompatibilities due to misidentification of the dam. Efforts must be made to improve the quality of pedigree information, with particular emphasis on information recorded at the calf's birth.

Effects of pedigree errors on the efficiency of conservation decisions

Conservation schemes often aim at increasing genetic diversity by minimizing kinship, and the best method to achieve this goal, when pedigree data is available, is to apply optimal contributions. Optimal contributions calculate contributions per animal so that the weighted average mean kinship among candidate parents is minimized. This approach assumes that pedigree data is correct and complete. However, in practice, pedigrees often contain errors: parents are recorded incorrectly or even missing. We used simulations to investigate the effect of these two types of errors on minimizing kinship. Our findings show that a low percentage of wrong parent information reduces the effect of optimal contributions. When the percentage of wrong parent information is above 15%, the population structure and type of errors, should be taken into account before applying optimal contributions. Optimal contributions based on pedigrees with missing parent information hampers conservation of genetic diversity; however, missing parent information can be corrected. It is crucial to know which animals are founders. We strongly recommend that pedigree registration include whether missing parents are either true founders or non-founders.

Udder health shows inbreeding depression in Danish Holsteins

The purpose of this study was to estimate inbreeding depression for udder health using records on mastitis incidence and somatic cell count (SCC). Data were selected based on a pedigree completeness index so that inbreeding coefficients were reliable. Average inbreeding was 3.3%. The presence or absence of mastitis was considered in first (CM1), second (CM2), and third (CM3) lactations, and was recorded as a binary trait. Somatic cell count was also included and analyzed on the log scale. A minimum of 140,000 cows with records were included per trait. Linear sire models were used and (co)variance components were estimated. Linear and curvilinear effects of the coefficient of inbreeding were included. Inbreeding significantly affected all traits investigated. Three traits, CM1, CM3, and SCC, showed a nonlinear relationship between phenotype and inbreeding coefficient. A cow with 5% inbreeding compared with a cow with 2% inbreeding had a higher SCC and a higher incidence of mastitis. The SCC increased by 1,500 cells/mL in first lactation and the incidence of mastitis increased by 1.08, 0.55, and 0.98% in first, second, and third lactation, respectively. The corresponding reduction in net return from these traits over 3 lactations amounted to 11.00 US dollars under Danish production circumstances. These results are a step on the way to estimating the total cost of inbreeding depression to determine an acceptable rate of inbreeding.

A framework for power and sensitivity analyses for quantitative genetic studies of natural populations, and case studies in Soay sheep (Ovis aries)

Studies of the quantitative genetics of natural populations have contributed greatly to evolutionary biology in recent years. However, while pedigree data required are often uncertain (i.e. incomplete and partly erroneous) and limited, means to evaluate the effects of such uncertainties have not been developed. We have therefore developed a general framework for power and sensitivity analyses of such studies. We propose that researchers first generate a set of pedigree data that they wish to use in a quantitative genetic study, as well as data regarding errors that occur in that pedigree. This pedigree is then permuted using the data regarding errors to generate hypothetical 'true' and 'assumed' pedigrees that differ so as to mimic pedigree errors that might occur in the study system under consideration. Phenotypic data are then simulated across the true pedigree (according to user-defined genetic and environmental covariance structures), before being analysed with standard quantitative genetic techniques in conjunction with the 'assumed' pedigree data. To illustrate this approach, we conducted power and sensitivity analyses in a well-known study of Soay sheep (Ovis aries). We found that, although the estimation of simple genetic (co)variance structures is fairly robust to pedigree errors, some potentially serious biases were detected under more complex scenarios involving maternal effects. Power analyses also showed that this study system provides high power to detect heritabilities as low as about 0.09. Given this range of results, we suggest that such power and sensitivity analyses could greatly complement empirical studies, and we provide the computer program PEDANTICS to aid in their application.

Technical note: a note on the differential impact of wrong and missing sire information on reliability and gain

This note analytically derives the impact that wrong and missing sire information (WSI and MSI, respectively) has on the reliability of predicting merit and gain compared with perfect information. In particular, for small WSI and MSI, WSI was shown to have twice the impact of MSI for both reliability and gain, and the impact of both WSI and MSI increased as the reliability of predicting merit with perfect information decreased. The overall impact on the efficiency of gain for small WSI and MSI was half the overall impact on reliability.

Wrong and missing sire information affects genetic gain in the Angeln dairy cattle population

In the present study, molecular genetic markers were used to help estimate the degree of wrong sire information in the German Angeln dairy cattle population. Sixteen polymorphic microsatellite markers were genotyped on 5 different paternal half-sib families with a total of 805 daughters. For the genotyping process, blood samples of the daughters and semen samples of the sires were used. Allelic frequencies and exclusion probabilities were estimated. The simultaneous effect of wrong (WSI) and missing sire information (MSI) on the reliability of estimated breeding values and on the genetic gain was investigated using deterministic simulations. For these simulations, different values for the number of daughters per sire, heritability, WSI, and MSI were chosen. The estimated proportion of the WSI was 7% in the German Angeln dairy cattle population. The combined impact of WSI and MSI on the genetic gain was relatively large, especially in the case of small progeny size per sire and low heritability. The impact of WSI was more harmful than MSI on response to selection.

Genetic evaluation using parentage information from genetic markers1

Genetic evaluation relies on pedigree information to account for the trait information on individuals and their relatives. Recording pedigrees may place unfavorable restrictions on the management of breeding populations, such as the use of single-sire mating groups and the observation of parturition. The use of DNA marker information is an alternative method to identify parents, but it is difficult to assign the parents unambiguously for all progeny in extensively farmed livestock without the use of very many markers. We present methods that use DNA information on parentage within a genetic evaluation system that allow for genotyping errors and for the parentage information to be incomplete, with probabilities assigned to possible parent pairs (i.e., fractional parentage assignment). Two of these methods use a computing strategy that circumvents the high memory requirements associated with the application of previous methods designed for use with fractional parentage assignment. This strategy has an additional advantage of allowing the same statistical models to be used in the evaluation as with recorded pedigrees. The use of DNA marker-based parentage for genetic evaluation is associated with lower genetic gain (at the same survival levels) than by using the true pedigree. This decrease in gain depends on a number of factors, including trait heritability and the DNA markers used. The methods we have described show how DNA marker information could be used to replace traditional pedigree recording.

Analysis of microsatellites and parentage testing in saltwater crocodiles

Fifteen microsatellite loci were evaluated in farmed saltwater crocodiles for use in parentage testing. One marker (C391) could not be amplied. For the remaining 14, the number of alleles per locus ranged from two to 16, and the observed heterozygosities ranged from 0.219 to 0.875. The cumulative exclusion probability for all 14 loci was .9988. the 11 loci that showed the greatest level of polymorphism were used for parentage testing, with an exclusion probability of .9980. With these 11 markers on 107 juveniles from 16 known breeding pairs, a 5.6% pedigree error rate was detected. This level of pedigree error, if consistent, could have an impact on the accuracy of gentic parameter and breeding value estimation. The usefulness of these markers was also evaluated for assigning parentage in situations where maternity, paternity, or both may not be known. In these situations, a 2% error in parentage assignment was predicted. It is therefore recommended that more micro-satellite markers be used in these situations. The use of these microsatellite markers will broaden the scope of a breeding program, allowing progeny to be tested from adults maintained in large breeding lagoons for selection as future breeding animals.

Factors affecting incorrect paternity assignment in the Israeli Holstein population

A total of 6040 Israeli Holstein cows from 181 Kibbutz herds listed as progeny of 11 sires were genotyped for 104 microsatellites. Seventeen markers were deleted due to a frequency of erroneous genotypes >1%, leaving 160,470 valid genotypes. Conflicts between the putative sire and daughter in at least 2 markers and for at least 10% of the markers genotyped per cow were required to reject paternity. Cows that did not meet the requirements for paternity confirmation or rejection were deleted from further analysis. The frequency of rejected paternity was 11.7%. The effects of recorded sire, birth year, geographical region, herd, and inseminator on the frequency of paternity rejection were analyzed with linear and nonlinear models. Only the effects of inseminator and recorded sire were significant in all models tested that included these effects. The main causes of incorrect paternity recording appear to be inseminator recording mistakes, and possibly mistakes with respect to semen labeling at the AI institutes. Incorrect paternity recording due to multiple inseminations by different sires could explain, at most, 20% of the paternity mistakes. Instituting a system of quality control, especially at the level of the inseminator, should reduce paternity errors to no more than 8%, and increase genetic progress by at least 1%.

Estimating genetic parameters in natural populations using the "animal model"

Estimating the genetic basis of quantitative traits can be tricky for wild populations in natural environments, as environmental variation frequently obscures the underlying evolutionary patterns. I review the recent application of restricted maximum-likelihood "animal models" to multigenerational data from natural populations, and show how the estimation of variance components and prediction of breeding values using these methods offer a powerful means of tackling the potentially confounding effects of environmental variation, as well as generating a wealth of new areas of investigation.

Discrimination between selected lines of pigs using AFLP® markers

Amplified fragment length polymorphic (AFLP) markers were used to discriminate between lines of pigs, divergently selected over seven generations for components of efficient lean growth rate. A total of 270 animals with 30 animals per line were genotyped for 239 polymorphic AFLP markers. Canonical variate analysis identified linear combinations of the AFLP marker scores that grouped animals by selection line with no overlap between selection lines. Cluster analysis of AFLP marker scores identified 10 groups of animals with 226 of the 270 animals clustered into nine groups, each consisting of animals from only one selection line. AFLP marker genotyping, using the EcoRI and TaqI restriction enzymes, provided an effective means of discriminating between animals of different selection lines that have arisen from one base population.