Background: PLINK is probably the most used program for analyzing SNP genotypes and runs of homozygosity (ROH), both in human and in animal populations. The last decade, ROH analyses have become the state-of-the-art method for inbreeding assessment. In PLINK, the-homozyg function is used to perform ROH analyses and relies on several input settings. These settings can have a large impact on the outcome and default values are not always appropriate for medium density SNP array data. Guidelines for a robust and uniform ROH analysis in PLINK using medium density data are lacking, albeit these guidelines are vital for comparing different ROH studies. In this study, 8 populations of different livestock and pet species are used to demonstrate the importance of PLINK input settings. Moreover, the effects of pruning SNPs for low minor allele frequencies and linkage disequilibrium on ROH detection are shown. Results: We introduce the genome coverage parameter to appropriately estimate F ROH and to check the validity of ROH analyses. The effect of pruning for linkage disequilibrium and low minor allele frequencies on ROH analyses is highly population dependent and such pruning may result in missed ROH. PLINK's minimal density requirement is crucial for medium density genotypes and if set too low, genome coverage of the ROH analysis is limited. Finally, we provide recommendations for the maximal gap, scanning window length and threshold settings. Conclusions: In this study, we present guidelines for an adequate and robust ROH analysis in PLINK on medium density SNP data. Furthermore, we advise to report parameter settings in publications, and to validate them prior to analysis. Moreover, we encourage authors to report genome coverage to reflect the ROH analysis' validity. Implementing these guidelines will substantially improve the overall quality and uniformity of ROH analyses.

Background Runs of homozygosity (ROH) have become the state-of-the-art method for analysis of inbreeding in animal populations. Moreover, ROH are suited to detect signatures of selection via ROH islands and are used in other applications, such as genomic prediction and genome-wide association studies (GWAS). Currently, a vast amount of single nucleotide polymorphism (SNP) data is available online, but most of these data have never been used for ROH analysis. Therefore, we performed a ROH analysis on large medium-density SNP datasets in eight animal species (cat, cattle, dog, goat, horse, pig, sheep and water buffalo; 442 different populations) and make these results publicly available. Results The results include an overview of ROH islands per population and a comparison of the incidence of these ROH islands among populations from the same species, which can assist researchers when studying other (livestock) populations or when looking for similar signatures of selection. We were able to confirm many known ROH islands, for example signatures of selection for the myostatin (MSTN) gene in sheep and horses. However, our results also included multiple other ROH islands, which are common to many populations and not identified to date (e.g. on chromosomes D4 and E2 in cats and on chromosome 6 in sheep). Conclusions We are confident that our repository of ROH islands is a valuable reference for future studies. The discovered ROH island regions represent a unique starting point for new studies or can be used as a reference for future studies. Furthermore, we encourage authors to add their population-specific ROH findings to our repository.

The Pi etrain pig originates from the Belgian village Pi etrain some time between 1920 and 1950. Owing to its superior conformation, the Pi etrain has spread worldwide since the 1960s. As initial population sizes were limited and close inbreeding was commonplace, the breed's genetic diversity has been questioned. Therefore, this study examines Pi etrain breed substructure, diversity and selection signatures using SNP data in comparison with Duroc, Landrace and Large White populations. Principal component analysis indicated three subpopulations, and F ST analysis showed that US Pi etrains differ most from European Pi etrains. Average inbreeding based on runs of homozygosity (ROH) segments larger than 4 Mb ranged between 16.7 and 20.9%. The highest chromosomal inbreeding levels were found on SSC8 (42.7%). ROH islands were found on SSC8, SSC15 and SSC18 in all Pi etrain populations, but numerous population-specific ROH islands were also detected. Moreover, a large ROH island on SSC8 (34-126 Mb) appears nearly fixed in all Pi etrain populations, with a unique genotype. Chromosomal ROH patterns were similar between Pi etrain populations. This study shows that Pi etrain populations are genetically diverging, with at least three genetically distinct populations worldwide. Increasing genetic diversity in local Pi etrain populations by introgression from other Pi etrain populations seems to be only limited. Moreover, a unique 90 Mb region on SSC8 appeared largely fixed in the Pi etrain breed, indicating that fixation was already present before the 1960s. We believe that strong selection and inbreeding during breed formation fixed these genomic regions in Pi etrains. Finally, we hypothesize that independent coat color selection may have led to large ROH pattern similarities on SSC8 between unrelated pig breeds.

Background Show jumping is one of the most popular disciplines in the horse sector, which makes success in show jumping competitions an important breeding goal for many studbooks. Therefore, the genetic evaluation of show jumping performance is of major interest and this is the case for two Belgian Warmblood studbooks: the Belgian Warmblood horse and Zangersheide. In this study, first an improved phenotype for show jumping performance was developed, i.e. adjusted fence height based on a new non-arbitrary method to scale ranking and competition level, which are two major components of success in competitions. Second, we assessed the importance of including a rider effect in genetic models for show jumping performance, this effect being under debate in sport horse breeding. Third, genetic models based on elementary performances and one model based on a summarized performance were compared in terms of model fit, heritabilities and the stability of estimated breeding values to define the most suitable one for the genetic evaluation of show jumping performance. Results In this study, more than 600,000 Belgian competition records and almost 81,000 horses were used. Genetic evaluations were developed based on elementary performances (Blom-transformed ranking and adjusted fence height) and on a summarized performance (highest level achieved). Estimated heritabilities of Blom-transformed ranking, adjusted fence height and highest level achieved were 0.09, 0.12 and 0.39, respectively. Including a rider effect improved the models for genetic evaluations. Estimated genetic correlations between the studied models were moderate to high (rg = 0.60–0.99). With the best fit model, the accuracy of the estimated breeding value (EBV) for adjusted fence height reached 0.70 for a larger number of stallions and for stallions that tended to be younger. Conclusions We recommend breeders to implement this new phenotype ‘adjusted fence height’ in breeding programs. It is moderately to highly correlated with Blom-transformed ranking and highest level achieved, a proxy for lifetime success, and is available for selection candidates from an early age onwards.

No abstract