A recent comprehensive genomic analysis based on 50K SNP profiles has shown that the regional Bal... more A recent comprehensive genomic analysis based on 50K SNP profiles has shown that the regional Balkan sheep populations have considerable genetic overlap but are distinctly different from surrounding breeds. All eight Croatian sheep breeds were represented by a small number of individuals per breed. Here, we genotyped 220 individuals representing the native Croatian sheep breeds (Istrian Sheep, Krk Island Sheep, Cres Island Sheep, Rab Island Sheep, Lika Pramenka, Pag Island Sheep, Dalmatian Pramenka, Dubrovnik Sheep) and mouflon using the Ovine Infinium® HD SNP BeadChip (606,006 SNPs). In addition, we included publicly available Balkan Pramenka and other Mediterranean sheep breeds. Our analyses revealed the complex population structure of Croatian sheep breeds and their origin and geographic barriers (island versus mainland). Migration patterns confirmed the historical establishment of breeds and the pathways of gene flow. Inbreeding coefficients (FROH>2 Mb) between sheep populati...
Proceedings of the World Congress on Genetics Applied to Livestock Production, 2018
Heterosis is a measure of the superior performance of the crossbreds relative to the average of p... more Heterosis is a measure of the superior performance of the crossbreds relative to the average of purebred parental populations involved in cross-breeding, due to favorable non-allelic gene interactions. Ignoring epistatic interactions, the extent of heterosis depends on the degree of dominance and its direction and difference in allele frequencies of the contributing loci. In this study we analyzed the possible association of heterosis with a sperm quality trait (percentage of live sperm) in bulls of admixed Swiss Fleckvieh, a composite of Simmental and Holstein Friesian, applying genome-wide mapping with genetic markers (SNP). Total of 41,749 phenotypic records of percentage of live sperm for both purebred and admixed bulls were used. After quality control of genotypes, 1,169 bulls with 38,205 SNP remained for analyses. The model for single locus mapping consisted of genetic effect of bulls, fixed effects (age, contemporary groups, ejaculate intervals and semen collector), additive SNP effect, genomic breed percent and genomic breed dominance component of heterosis. For percentage of live sperm 10 significant signals on chromosomes 3, 4, 5, 7, 13 and 14 were detected. Four of these regions contained genes related to spermatogenesis.
Journal of Animal Breeding and Genetics, Mar 30, 2015
GRain is freely available software intended to enable and promote testing of hypotheses with resp... more GRain is freely available software intended to enable and promote testing of hypotheses with respect to purging and heterogeneity of inbreeding depression. The program is based on a stochastic approach, the gene dropping method, and calculates various coefficients from large and complex pedigrees. GRain calculates, together with the 'classical' inbreeding coefficient, ancestral inbreeding coefficients proposed by Ballou, (1997) J. Hered., 88, 169 and Kalinowski et al., (2000) Conserv. Biol., 14, 1375 as well as an ancestral history coefficient (A HC), defined here for the first time. A HC is defined as the number that tells how many times during pedigree segregation (gene dropping) a randomly taken allele has been in IBD status. Furthermore, GRain enables testing of heterogeneity and/or purging of inbreeding depression with respect to different founders/ancestors by calculating partial coefficients for all previously obtained coefficients.
The objective of this study was to develop a DNA test for rapid characterisation of goat k-casein... more The objective of this study was to develop a DNA test for rapid characterisation of goat k-casein (k-CN) A and B variants and to study this polymorphism in Italian goat breeds. Genetic polymorphism of k-CN gene was, in addition to isoelectric focusing, analysed according to a new technique designated as amplification created restriction site. Two alleles that differ in one nucleotide mutation (G AE A) in exon 4 were characterised. The 167-bp PCR product surrounding the nucleotide mutation was amplified from genomic DNA and the PCR product was digested with MaeIII. After digestion the A allele gives three fragments of 77, 65 and 25 bp in comparison with the B allele which gives two fragments of 90 and 77 bp. The analysis of allele frequency distribution at k-CN locus, based on 401 individual samples, revealed significant differences among three goat breeds from the north of Italy (Nera di Verzasca, Frontalasca and Alpine) with frequency of k-CN B allele around 0.3, versus two goat breeds from the south of Italy (Maltese and Sarda) with frequency of k-CN B allele around 0.5. While two goat breeds (Maltese and Nera di Verzasca) did not show significant deviations from the Hardy-Weinberg equilibrium, a highly significant excess of heterozygote genotype (AB) was observed in Alpine, Frontalasca and Sarda goats. Here the developed DNA method and observed relatively high frequency of k-CN B allele give a prerequisite for the assessment of research related to the simultaneous estimation of the effects of composite a s1 /k-CN genotypes on milk production and cheese-making properties.
The aim of this study was to estimate genetic parameters for coat color in horses. Besides defini... more The aim of this study was to estimate genetic parameters for coat color in horses. Besides defining coat color classes (gray, chestnut, bay, and black), the phenotypes were also measured quantitatively according to standardized international procedures (Commission Internationale de l'Eclairage L*, a*, b*), where L* describes lightness, a* describes color saturation from red to green, and b* describes color saturation from yellow to blue. The total color saturation was derived from a* and b* and referred to as Chroma. A total of 294 horses from the breeds Lipizzan, Nonius, Arabian Pure Bred, Shagya Arabian, and Gidran were measured at neck, shoulder, and belly. Heritabilities (within and between breeds or color classes) and repeatabilities were estimated using REML from univariate animal models defined separately for gray and nongray horses. For gray horses, the estimated within-breed heritabilities for L* ranged from 0.45 to 0.49 and for a*, b*, and Chroma from 0.09 to 0.52, indicating moderate polygenic effect. For nongray horses, between-color class heritabilities were high (0.70 to 0.85) and within-color class heritabilities were negligible (except for L* measured on neck and belly, 0.21 and 0.34, respectively). Additionally, the importance of L* was described by the relation with the total melanin content of horse coat hair; for gray and nongray horses, a strong negative linear relationship was detected (P < 0.01). The spectrometric measures and the results of this study demonstrate a possible approach to the estimation of the polygenic component involved in coat color inheritance.
A genome-wide scan of Slovak Pinzgau cattle was prepared for the first time in order to estimate ... more A genome-wide scan of Slovak Pinzgau cattle was prepared for the first time in order to estimate their genetic diversity at a more detailed level compared to previously published studies. The aim of this study was to describe the genetic diversity based on the runs of homozygosity (ROH s), linkage disequilibrium (LD) and effective population size (N eLD) using genome-wide data. Moreover, Bayesian clustering algorithms and multivariate methods were used to detect the population structure, potential admixture level and relationship between Austrian and Slovak Pinzgau cattle with respect to a large meta-population consisting of 15 European cattle breeds. The proportion of ROH segments ranged from 0.43 to 1.91% in Slovak Pinzgau, depending on the minimum size of an ROH. The genomic inbreeding coefficients were higher than the pedigree ones possibly due to the limited number of available generations in pedigree data. The observed N eLD was close to the limit value characterizing the endangerment Electronic supplementary material The online version of this article (
Agriculturae Conspectus Scientificus (Poljoprivredna Znanstvena Smotra), Jan 12, 2000
The relationship between inbreeding and melanoma status (graded from 0 to 4) was analysed by vari... more The relationship between inbreeding and melanoma status (graded from 0 to 4) was analysed by various regression models. Analysed data referred to 296 grey Lipizzan horses originating from five state-owned studs (Austria, Croatia, Hungary, Slovakia and Slovenia) and with average inbreeding coefficient (F=0.107) calculated from extremely informative pedigrees (98% and 76% of horses had completely full pedigree in generation 10 and 20, respectively). In all regression models, in addition, the effects of stud (fixed) and age (covariate) were included. When all data were treated as one population, the estimates from linear and ancestral inbreeding models were not significant. Total inbreeding effect estimates (at F=0.125 and Fa=0.57) were 0.26 and 0.30 for the ancestral inbreeding and linear regression models, respectively. Heterogeneity among state-owned studs in inbreeding effects was also tested for both models and weak statistical significance was obtained for the interaction model with ancestral inbreeding (P=0.049). However, observed effect in the model with interaction was not consistent, did not yield in better model fitting and the obtained significance is probably just a statistical artefact. In general, although some indications about the relationship between ancestral inbreeding and melanoma were present, inbreeding does not appear to be a factor that substantially influences the expression of melanoma in Lipizzan horses.
Background: Genome-wide prediction has become the method of choice in animal and plant breeding. ... more Background: Genome-wide prediction has become the method of choice in animal and plant breeding. Prediction of breeding values and phenotypes are routinely performed using large genomic data sets with number of markers on the order of several thousands to millions. The number of evaluated individuals is usually smaller which results in problems where model sparsity is of major concern. The LASSO technique has proven to be very well-suited for sparse problems often providing excellent prediction accuracy. Several computationally efficient LASSO algorithms have been developed, but optimization of hyper-parameters can be demanding. Results: We have developed a novel automatic adaptive LASSO (AUTALASSO) based on the alternating direction method of multipliers (ADMM) optimization algorithm. The two major hyper-parameters of ADMM are the learning rate and the regularization factor. The learning rate is automatically tuned with line search and the regularization factor optimized using Golden section search. Results show that AUTALASSO provides superior prediction accuracy when evaluated on simulated and real bull data compared to the adaptive LASSO, LASSO and ridge regression implemented in the popular glmnet software. Conclusions: The AUTALASSO provides a very flexible and computationally efficient approach to GWP, especially when it is important to obtain high prediction accuracy and genetic gain. The AUTALASSO also has the capability to perform GWAS of both additive and dominance effects with smaller prediction error than the ordinary LASSO.
All over Europe, the number of draught horses has decreased drastically during the last 50 years.... more All over Europe, the number of draught horses has decreased drastically during the last 50 years. As a prerequisite for efficient management decisions, we analysed the conservation status in Austrian (Noriker Carinthia-NC, Noriker Salzburg-NS), Croatian (Croatian Coldblood-C, Posavina horse-P) and German (Altmaerkisch Coldblood-A, Black Forest horse-BF, Mecklenburg Coldblood-M, Rhenish German Draught horse-R, Saxon Thuringa Coldblood-ST, Schleswig Draught horse-Sch, South German Coldblood-SG) draught horses (434) using multilocus genotypic information from 30 (effectively 27) microsatellite loci. Populations located in areas with less intensive agricultural production (C, NC, NS, P and SG) had greater diversity within the population and estimated effective population size than A, BF, Sch, M, R and ST populations. The PCA plots revealed that populations form five separate groups. The 'Noriker' group (NC, NS and SG) and the 'Rhenish' group (A, M, R and ST) were the most distinctive (pairwise F ST values ranged from 0.078 to 0.094). The 'Croatian' group (C and P) was in the centre, while the BF and Sch populations formed two out-groups. A posterior Bayesian analysis detected further differentiation, mainly caused by political and geographical factors. Thus, it was possible to separate the South German Coldblood from the Austrian Noriker population where no subpopulation structure was detected. The admixture analysis revealed imprecise classification between C and P populations. A small but notable separation of R from A, M and ST populations was detected, while Sch and BF populations remained as out-groups. The information obtained should aid in making efficient conservation decisions.
Estimates of the inbreeding coefficient based on runs of homozygosity (F ROH) are widely used in ... more Estimates of the inbreeding coefficient based on runs of homozygosity (F ROH) are widely used in population genetics and livestock studies. The accuracy of F ROH estimates is determined in part by the density of the genotyping chip. Compared with high density (HD) chips, low density (LD) chips tend to overestimate F ROH in shorter homozygous segments. It is unclear whether imputation of genotypes from LD to HD would reduce or eliminate this overestimation. Our goal was to determine whether F ROH calculations based on imputed LD genotypes are accurate. To this end, F ROH estimates were determined for Nelore cows using HD or LD imputed genotypes (F ROH_HD and F ROH_LDimp). Our results showed that LD imputed genotypes reduced but did not eliminate differences between estimates. Differences were observed in both directions, with under-and overestimates of F ROH. This suggests that imputation errors can either disrupt or generate homozygous segments.
Variation in mitochondrial DNA (mtDNA) and Y-chromosome haplotypes was analysed in nine domestic ... more Variation in mitochondrial DNA (mtDNA) and Y-chromosome haplotypes was analysed in nine domestic sheep breeds (159 rams) and 21 mouflon (Ovis musimon) sampled in the East Adriatic. Mitochondrial DNA analyses revealed a high frequency of type B haplotypes, predominantly in European breeds, and a very low frequency of type A haplotypes, which are more frequent in some Asian breeds. Mitochondrial haplotype Hmt-3 was the most frequent (26.4%), and 37.1%, 20.8% and 7.6% of rams had haplotypes one, two and three mutations remote from Hmt-3 respectively. In contrast, Y-chromosome analyses revealed extraordinary paternal allelic richness: HY-6, 89.3%; HY-8, 5.0%; HY-18, 3.1%; HY-7, 1.3%; and HY-5, 1.3%. In fact, the number of haplotypes observed is comparable to the number found in Turkish breeds and greater than the number found in European breeds so far. Haplotype HY-18 (A-oY1/135-SRYM18), identified here for the first time, provides a link between the haplotype HY-12 (A-oY1/139-SRYM18) found in a few rams in Turkey and haplotype HY-9 (A-oY1/131-SRYM18) found in one ram in Ethiopia. All mouflons had type B mtDNA haplotypes, including the private haplotype (Hmt-55), and all were paternally monomorphic for haplotype HY-6. Our data support a quite homogeneous maternal origin of East Adriatic sheep, which is a characteristic of European breeds. At the same time, the high number of haplotypes found was surprising and intriguing, and it begs for further analysis. Simultaneous analysis of mtDNA and Y-chromosome information allowed us to detect a large discrepancy between maternal and paternal lineages in some populations. This is most likely the result of breeder efforts to 'upgrade' local populations using rams with different paternal origins.
The aim of this study was to predict individual age of admixture in the crossbred Swiss fleckvieh... more The aim of this study was to predict individual age of admixture in the crossbred Swiss fleckvieh population. We checked how well the method is dealing with recent admixture with high throughput single nucleotide polymorphism data from the bovine 50K SnP Chip. A total of 101 red holstein, 91 Simmental, and 308 crossed animals were available for analysis. Age of admixture was derived from the complete pedigree and molecular markers. The method applied (using SABEr software) based on markov-hidden markov model was able to derive age of admixture similar to estimates of pedigree data, however the values were often overestimated. Of 21 investigated cases, results from SnP data reflected paternal and maternal age of admixture well for 9 cases but provided results out of range for the other 12 cases. Alternative methods based on breed-specific haplotype blocks need to be evaluated in the future.
Abstract: On the Croatian part of Adriatic Sea (coastal part and islands), currently is being bre... more Abstract: On the Croatian part of Adriatic Sea (coastal part and islands), currently is being bred more that 300 000 sheep. Despite its regional specificities and variability, detailed genetic analysis of autochthonous sheep populations in Croatia has been not carried out ...
Estimation of variance components for a quantitative trait is very difficult when genetic interac... more Estimation of variance components for a quantitative trait is very difficult when genetic interactions between loci can not be neglected. The presence of interactions, even among few loci, very soon becomes extremely complicated due to the large number of possible combinations. Here, we presented development of variances components in the simplest model that includes epistatic effects (two locus bi-allelic model) for a population that is in Hardy-Weinberg equilibrium. The presented formulas enable three dimension graphical presentations of all variance components (including additive by additive, additive by dominance and dominance by dominance variances) under different models for all possible gene frequencies and can be used in scholastic purposes. Further, these formulas are also suitable for Monte Carlo based studies related to deviations of the observed from theoretical values (for example the effects of inbreeding on epistatic variances) as well as for further developments in segregation analyses related to two loci models and epistatic effects.
The spread of farming into the Northern Mediterranean was frequently analysed through maternal an... more The spread of farming into the Northern Mediterranean was frequently analysed through maternal and paternal DNA lineages of humans and domestic animals, as an important topic for understanding the Neolitic agricultural revolution. We analyzed mitochondrial DNA (partly control region and 12S rRNA coding region and fully tRNAPhe) and specifi c Y chromosome (microsatellite SRYM18 and SNP AY604734.2:g.67A>G) polymorphisms of 159 domestic sheep (9 breeds) and 21 moufl on (Ovis musimon) rams sampled from Eastern Adriatic, all along Croatian cost. Maternal and paternal lineages for, both, sheep and moufl on rams indicated ancestral homogeneity, with extensive variation around most frequent haplotypes. We indentifi ed 70 mtDNA haplotypes, including one new present only in moufl ons, and fi ve Y chromosome haplotypes, including H18 here described for the fi rst time (SRYM18 with 135). The large number of sheep (96.9%) and moufl on (100%) rams had haplotypes classifi ed in mtDNA haplogroup B. More precisely, there was 26.4% of sheep rams with H3 haplotype while 37.1%, 20.8% and 7.6% of rams had haplotypes that were one, two and three mutations remote, respectively. The remaining fi ve sheep rams had haplotypes classifi ed in mtDNA haplogroup A. Similarly, 89.3% of sheep and 100% of moufl on rams had Y chromosome haplotype H6 while seven, fi ve, three and two rams had haplotypes H8, H18, H7 and H5, respectively. Our study shows that present sheep populations of Eastern Adriatic, with exception of several rams that are indicating remote origin, are homogenous and belong to the lineages characteristic for European sheep populations.
Equine melanomas occur most commonly in grey horses at age 5 years or more. Generally, benign and... more Equine melanomas occur most commonly in grey horses at age 5 years or more. Generally, benign and malignant melanomas are distinguished by microscopy, but a more distinct classification would be helpful. The objectives of this study were to gain further evidence concerning the occurrence of melanotic tumours, and to evaluate the impact of heredity on melanoma development. A clinical study was conducted on a defined population of 296 grey horses of Lipizzaner breed. Individuals were classified according to their stage of disease using a 0-5 scale. Heritability was estimated on a sample of 296 grey horses with pedigrees traced back as far as 32 generations. Of the 296 horses, dermal melanomas were present in 148 horses (50%), 68 of which were more than age 15 years; 51 of these were melanoma-bearing. In 75.6% of cases, melanotic tumours were detected underneath the tail. Although melanoma-bearing grey horses were encountered up to stage 4, none of the affected individuals suffered any severe clinical effect or was handicapped in performance. Statistical analysis revealed highly significant effects of stud and age (P < 0.0001), explaining 28% of the total variability. In contrast to melanomas in solid-coloured horses characterised by early metastases, melanomas in grey horses showed less malignancy. Affected individuals often had encapsulated nodules or structures similar to human blue nevi. Grey horse-specific genetic factors inhibiting metastatic processes may be responsible for this phenomenon. Although the obtained heritability estimate of 0.36 with a standard error of 0.11 indicates a strong genetic impact on the development of melanoma in ageing grey horses, a possible influence of the genes with large effects was also suggested. Therefore, further analysis is required of melanoma development in the ageing grey horse.
It is well known that cheese yield and quality are affected by animal genetics, milk quality (che... more It is well known that cheese yield and quality are affected by animal genetics, milk quality (chemical, physical, and microbiological), production technology, and the type of rennet and dairy cultures used in production. Major differences in the same type of cheese (i.e., hard cheese) are caused by the rennet and dairy cultures, which affect the ripening process. This review aims to explore current technological advancements in animal genetics, methods for the isolation and production of rennet and dairy cultures, along with possible applications of microencapsulation in rennet and dairy culture production, as well as the challenge posed to current dairy technologies by the preservation of biodiversity. Based on the reviewed scientific literature, it can be concluded that innovative approaches and the described techniques can significantly improve cheese production.
Proceedings of 12th World Congress on Genetics Applied to Livestock Production (WCGALP)
Livestock Guardian Dogs (LGDs, Canis familiaris) are large dogs adapted to harsh mountain environ... more Livestock Guardian Dogs (LGDs, Canis familiaris) are large dogs adapted to harsh mountain environments and selected to protect grazing livestock from predators in free-ranging grazing systems. Signals of positive selection in a meta-population represented by nine LGD breeds (116 individuals) were analysed based on information from 104,618 SNP genotypes using the extreme Runs of Homozygosity islands (eROHi) and the integrated Haplotype Score (iHS) approach. We found that 28 genomic regions (11 identified using the eROHi approach and 17 using the iHS approach) had highly significant signals for positive selection. Some of the most pronounced selection signals are associated with genes such as LRIG3, TRIP11, CNTNAP2 and EFNA5. The presented results will contribute to the understanding of the synergistic behavioural interaction of LGDs with humans and grazing animals, especially sheep and goats, and the mammals' adaptation to harsh mountain conditions, as well as to their breeding management.
A recent comprehensive genomic analysis based on 50K SNP profiles has shown that the regional Bal... more A recent comprehensive genomic analysis based on 50K SNP profiles has shown that the regional Balkan sheep populations have considerable genetic overlap but are distinctly different from surrounding breeds. All eight Croatian sheep breeds were represented by a small number of individuals per breed. Here, we genotyped 220 individuals representing the native Croatian sheep breeds (Istrian Sheep, Krk Island Sheep, Cres Island Sheep, Rab Island Sheep, Lika Pramenka, Pag Island Sheep, Dalmatian Pramenka, Dubrovnik Sheep) and mouflon using the Ovine Infinium® HD SNP BeadChip (606,006 SNPs). In addition, we included publicly available Balkan Pramenka and other Mediterranean sheep breeds. Our analyses revealed the complex population structure of Croatian sheep breeds and their origin and geographic barriers (island versus mainland). Migration patterns confirmed the historical establishment of breeds and the pathways of gene flow. Inbreeding coefficients (FROH>2 Mb) between sheep populati...
Proceedings of the World Congress on Genetics Applied to Livestock Production, 2018
Heterosis is a measure of the superior performance of the crossbreds relative to the average of p... more Heterosis is a measure of the superior performance of the crossbreds relative to the average of purebred parental populations involved in cross-breeding, due to favorable non-allelic gene interactions. Ignoring epistatic interactions, the extent of heterosis depends on the degree of dominance and its direction and difference in allele frequencies of the contributing loci. In this study we analyzed the possible association of heterosis with a sperm quality trait (percentage of live sperm) in bulls of admixed Swiss Fleckvieh, a composite of Simmental and Holstein Friesian, applying genome-wide mapping with genetic markers (SNP). Total of 41,749 phenotypic records of percentage of live sperm for both purebred and admixed bulls were used. After quality control of genotypes, 1,169 bulls with 38,205 SNP remained for analyses. The model for single locus mapping consisted of genetic effect of bulls, fixed effects (age, contemporary groups, ejaculate intervals and semen collector), additive SNP effect, genomic breed percent and genomic breed dominance component of heterosis. For percentage of live sperm 10 significant signals on chromosomes 3, 4, 5, 7, 13 and 14 were detected. Four of these regions contained genes related to spermatogenesis.
Journal of Animal Breeding and Genetics, Mar 30, 2015
GRain is freely available software intended to enable and promote testing of hypotheses with resp... more GRain is freely available software intended to enable and promote testing of hypotheses with respect to purging and heterogeneity of inbreeding depression. The program is based on a stochastic approach, the gene dropping method, and calculates various coefficients from large and complex pedigrees. GRain calculates, together with the 'classical' inbreeding coefficient, ancestral inbreeding coefficients proposed by Ballou, (1997) J. Hered., 88, 169 and Kalinowski et al., (2000) Conserv. Biol., 14, 1375 as well as an ancestral history coefficient (A HC), defined here for the first time. A HC is defined as the number that tells how many times during pedigree segregation (gene dropping) a randomly taken allele has been in IBD status. Furthermore, GRain enables testing of heterogeneity and/or purging of inbreeding depression with respect to different founders/ancestors by calculating partial coefficients for all previously obtained coefficients.
The objective of this study was to develop a DNA test for rapid characterisation of goat k-casein... more The objective of this study was to develop a DNA test for rapid characterisation of goat k-casein (k-CN) A and B variants and to study this polymorphism in Italian goat breeds. Genetic polymorphism of k-CN gene was, in addition to isoelectric focusing, analysed according to a new technique designated as amplification created restriction site. Two alleles that differ in one nucleotide mutation (G AE A) in exon 4 were characterised. The 167-bp PCR product surrounding the nucleotide mutation was amplified from genomic DNA and the PCR product was digested with MaeIII. After digestion the A allele gives three fragments of 77, 65 and 25 bp in comparison with the B allele which gives two fragments of 90 and 77 bp. The analysis of allele frequency distribution at k-CN locus, based on 401 individual samples, revealed significant differences among three goat breeds from the north of Italy (Nera di Verzasca, Frontalasca and Alpine) with frequency of k-CN B allele around 0.3, versus two goat breeds from the south of Italy (Maltese and Sarda) with frequency of k-CN B allele around 0.5. While two goat breeds (Maltese and Nera di Verzasca) did not show significant deviations from the Hardy-Weinberg equilibrium, a highly significant excess of heterozygote genotype (AB) was observed in Alpine, Frontalasca and Sarda goats. Here the developed DNA method and observed relatively high frequency of k-CN B allele give a prerequisite for the assessment of research related to the simultaneous estimation of the effects of composite a s1 /k-CN genotypes on milk production and cheese-making properties.
The aim of this study was to estimate genetic parameters for coat color in horses. Besides defini... more The aim of this study was to estimate genetic parameters for coat color in horses. Besides defining coat color classes (gray, chestnut, bay, and black), the phenotypes were also measured quantitatively according to standardized international procedures (Commission Internationale de l'Eclairage L*, a*, b*), where L* describes lightness, a* describes color saturation from red to green, and b* describes color saturation from yellow to blue. The total color saturation was derived from a* and b* and referred to as Chroma. A total of 294 horses from the breeds Lipizzan, Nonius, Arabian Pure Bred, Shagya Arabian, and Gidran were measured at neck, shoulder, and belly. Heritabilities (within and between breeds or color classes) and repeatabilities were estimated using REML from univariate animal models defined separately for gray and nongray horses. For gray horses, the estimated within-breed heritabilities for L* ranged from 0.45 to 0.49 and for a*, b*, and Chroma from 0.09 to 0.52, indicating moderate polygenic effect. For nongray horses, between-color class heritabilities were high (0.70 to 0.85) and within-color class heritabilities were negligible (except for L* measured on neck and belly, 0.21 and 0.34, respectively). Additionally, the importance of L* was described by the relation with the total melanin content of horse coat hair; for gray and nongray horses, a strong negative linear relationship was detected (P < 0.01). The spectrometric measures and the results of this study demonstrate a possible approach to the estimation of the polygenic component involved in coat color inheritance.
A genome-wide scan of Slovak Pinzgau cattle was prepared for the first time in order to estimate ... more A genome-wide scan of Slovak Pinzgau cattle was prepared for the first time in order to estimate their genetic diversity at a more detailed level compared to previously published studies. The aim of this study was to describe the genetic diversity based on the runs of homozygosity (ROH s), linkage disequilibrium (LD) and effective population size (N eLD) using genome-wide data. Moreover, Bayesian clustering algorithms and multivariate methods were used to detect the population structure, potential admixture level and relationship between Austrian and Slovak Pinzgau cattle with respect to a large meta-population consisting of 15 European cattle breeds. The proportion of ROH segments ranged from 0.43 to 1.91% in Slovak Pinzgau, depending on the minimum size of an ROH. The genomic inbreeding coefficients were higher than the pedigree ones possibly due to the limited number of available generations in pedigree data. The observed N eLD was close to the limit value characterizing the endangerment Electronic supplementary material The online version of this article (
Agriculturae Conspectus Scientificus (Poljoprivredna Znanstvena Smotra), Jan 12, 2000
The relationship between inbreeding and melanoma status (graded from 0 to 4) was analysed by vari... more The relationship between inbreeding and melanoma status (graded from 0 to 4) was analysed by various regression models. Analysed data referred to 296 grey Lipizzan horses originating from five state-owned studs (Austria, Croatia, Hungary, Slovakia and Slovenia) and with average inbreeding coefficient (F=0.107) calculated from extremely informative pedigrees (98% and 76% of horses had completely full pedigree in generation 10 and 20, respectively). In all regression models, in addition, the effects of stud (fixed) and age (covariate) were included. When all data were treated as one population, the estimates from linear and ancestral inbreeding models were not significant. Total inbreeding effect estimates (at F=0.125 and Fa=0.57) were 0.26 and 0.30 for the ancestral inbreeding and linear regression models, respectively. Heterogeneity among state-owned studs in inbreeding effects was also tested for both models and weak statistical significance was obtained for the interaction model with ancestral inbreeding (P=0.049). However, observed effect in the model with interaction was not consistent, did not yield in better model fitting and the obtained significance is probably just a statistical artefact. In general, although some indications about the relationship between ancestral inbreeding and melanoma were present, inbreeding does not appear to be a factor that substantially influences the expression of melanoma in Lipizzan horses.
Background: Genome-wide prediction has become the method of choice in animal and plant breeding. ... more Background: Genome-wide prediction has become the method of choice in animal and plant breeding. Prediction of breeding values and phenotypes are routinely performed using large genomic data sets with number of markers on the order of several thousands to millions. The number of evaluated individuals is usually smaller which results in problems where model sparsity is of major concern. The LASSO technique has proven to be very well-suited for sparse problems often providing excellent prediction accuracy. Several computationally efficient LASSO algorithms have been developed, but optimization of hyper-parameters can be demanding. Results: We have developed a novel automatic adaptive LASSO (AUTALASSO) based on the alternating direction method of multipliers (ADMM) optimization algorithm. The two major hyper-parameters of ADMM are the learning rate and the regularization factor. The learning rate is automatically tuned with line search and the regularization factor optimized using Golden section search. Results show that AUTALASSO provides superior prediction accuracy when evaluated on simulated and real bull data compared to the adaptive LASSO, LASSO and ridge regression implemented in the popular glmnet software. Conclusions: The AUTALASSO provides a very flexible and computationally efficient approach to GWP, especially when it is important to obtain high prediction accuracy and genetic gain. The AUTALASSO also has the capability to perform GWAS of both additive and dominance effects with smaller prediction error than the ordinary LASSO.
All over Europe, the number of draught horses has decreased drastically during the last 50 years.... more All over Europe, the number of draught horses has decreased drastically during the last 50 years. As a prerequisite for efficient management decisions, we analysed the conservation status in Austrian (Noriker Carinthia-NC, Noriker Salzburg-NS), Croatian (Croatian Coldblood-C, Posavina horse-P) and German (Altmaerkisch Coldblood-A, Black Forest horse-BF, Mecklenburg Coldblood-M, Rhenish German Draught horse-R, Saxon Thuringa Coldblood-ST, Schleswig Draught horse-Sch, South German Coldblood-SG) draught horses (434) using multilocus genotypic information from 30 (effectively 27) microsatellite loci. Populations located in areas with less intensive agricultural production (C, NC, NS, P and SG) had greater diversity within the population and estimated effective population size than A, BF, Sch, M, R and ST populations. The PCA plots revealed that populations form five separate groups. The 'Noriker' group (NC, NS and SG) and the 'Rhenish' group (A, M, R and ST) were the most distinctive (pairwise F ST values ranged from 0.078 to 0.094). The 'Croatian' group (C and P) was in the centre, while the BF and Sch populations formed two out-groups. A posterior Bayesian analysis detected further differentiation, mainly caused by political and geographical factors. Thus, it was possible to separate the South German Coldblood from the Austrian Noriker population where no subpopulation structure was detected. The admixture analysis revealed imprecise classification between C and P populations. A small but notable separation of R from A, M and ST populations was detected, while Sch and BF populations remained as out-groups. The information obtained should aid in making efficient conservation decisions.
Estimates of the inbreeding coefficient based on runs of homozygosity (F ROH) are widely used in ... more Estimates of the inbreeding coefficient based on runs of homozygosity (F ROH) are widely used in population genetics and livestock studies. The accuracy of F ROH estimates is determined in part by the density of the genotyping chip. Compared with high density (HD) chips, low density (LD) chips tend to overestimate F ROH in shorter homozygous segments. It is unclear whether imputation of genotypes from LD to HD would reduce or eliminate this overestimation. Our goal was to determine whether F ROH calculations based on imputed LD genotypes are accurate. To this end, F ROH estimates were determined for Nelore cows using HD or LD imputed genotypes (F ROH_HD and F ROH_LDimp). Our results showed that LD imputed genotypes reduced but did not eliminate differences between estimates. Differences were observed in both directions, with under-and overestimates of F ROH. This suggests that imputation errors can either disrupt or generate homozygous segments.
Variation in mitochondrial DNA (mtDNA) and Y-chromosome haplotypes was analysed in nine domestic ... more Variation in mitochondrial DNA (mtDNA) and Y-chromosome haplotypes was analysed in nine domestic sheep breeds (159 rams) and 21 mouflon (Ovis musimon) sampled in the East Adriatic. Mitochondrial DNA analyses revealed a high frequency of type B haplotypes, predominantly in European breeds, and a very low frequency of type A haplotypes, which are more frequent in some Asian breeds. Mitochondrial haplotype Hmt-3 was the most frequent (26.4%), and 37.1%, 20.8% and 7.6% of rams had haplotypes one, two and three mutations remote from Hmt-3 respectively. In contrast, Y-chromosome analyses revealed extraordinary paternal allelic richness: HY-6, 89.3%; HY-8, 5.0%; HY-18, 3.1%; HY-7, 1.3%; and HY-5, 1.3%. In fact, the number of haplotypes observed is comparable to the number found in Turkish breeds and greater than the number found in European breeds so far. Haplotype HY-18 (A-oY1/135-SRYM18), identified here for the first time, provides a link between the haplotype HY-12 (A-oY1/139-SRYM18) found in a few rams in Turkey and haplotype HY-9 (A-oY1/131-SRYM18) found in one ram in Ethiopia. All mouflons had type B mtDNA haplotypes, including the private haplotype (Hmt-55), and all were paternally monomorphic for haplotype HY-6. Our data support a quite homogeneous maternal origin of East Adriatic sheep, which is a characteristic of European breeds. At the same time, the high number of haplotypes found was surprising and intriguing, and it begs for further analysis. Simultaneous analysis of mtDNA and Y-chromosome information allowed us to detect a large discrepancy between maternal and paternal lineages in some populations. This is most likely the result of breeder efforts to 'upgrade' local populations using rams with different paternal origins.
The aim of this study was to predict individual age of admixture in the crossbred Swiss fleckvieh... more The aim of this study was to predict individual age of admixture in the crossbred Swiss fleckvieh population. We checked how well the method is dealing with recent admixture with high throughput single nucleotide polymorphism data from the bovine 50K SnP Chip. A total of 101 red holstein, 91 Simmental, and 308 crossed animals were available for analysis. Age of admixture was derived from the complete pedigree and molecular markers. The method applied (using SABEr software) based on markov-hidden markov model was able to derive age of admixture similar to estimates of pedigree data, however the values were often overestimated. Of 21 investigated cases, results from SnP data reflected paternal and maternal age of admixture well for 9 cases but provided results out of range for the other 12 cases. Alternative methods based on breed-specific haplotype blocks need to be evaluated in the future.
Abstract: On the Croatian part of Adriatic Sea (coastal part and islands), currently is being bre... more Abstract: On the Croatian part of Adriatic Sea (coastal part and islands), currently is being bred more that 300 000 sheep. Despite its regional specificities and variability, detailed genetic analysis of autochthonous sheep populations in Croatia has been not carried out ...
Estimation of variance components for a quantitative trait is very difficult when genetic interac... more Estimation of variance components for a quantitative trait is very difficult when genetic interactions between loci can not be neglected. The presence of interactions, even among few loci, very soon becomes extremely complicated due to the large number of possible combinations. Here, we presented development of variances components in the simplest model that includes epistatic effects (two locus bi-allelic model) for a population that is in Hardy-Weinberg equilibrium. The presented formulas enable three dimension graphical presentations of all variance components (including additive by additive, additive by dominance and dominance by dominance variances) under different models for all possible gene frequencies and can be used in scholastic purposes. Further, these formulas are also suitable for Monte Carlo based studies related to deviations of the observed from theoretical values (for example the effects of inbreeding on epistatic variances) as well as for further developments in segregation analyses related to two loci models and epistatic effects.
The spread of farming into the Northern Mediterranean was frequently analysed through maternal an... more The spread of farming into the Northern Mediterranean was frequently analysed through maternal and paternal DNA lineages of humans and domestic animals, as an important topic for understanding the Neolitic agricultural revolution. We analyzed mitochondrial DNA (partly control region and 12S rRNA coding region and fully tRNAPhe) and specifi c Y chromosome (microsatellite SRYM18 and SNP AY604734.2:g.67A>G) polymorphisms of 159 domestic sheep (9 breeds) and 21 moufl on (Ovis musimon) rams sampled from Eastern Adriatic, all along Croatian cost. Maternal and paternal lineages for, both, sheep and moufl on rams indicated ancestral homogeneity, with extensive variation around most frequent haplotypes. We indentifi ed 70 mtDNA haplotypes, including one new present only in moufl ons, and fi ve Y chromosome haplotypes, including H18 here described for the fi rst time (SRYM18 with 135). The large number of sheep (96.9%) and moufl on (100%) rams had haplotypes classifi ed in mtDNA haplogroup B. More precisely, there was 26.4% of sheep rams with H3 haplotype while 37.1%, 20.8% and 7.6% of rams had haplotypes that were one, two and three mutations remote, respectively. The remaining fi ve sheep rams had haplotypes classifi ed in mtDNA haplogroup A. Similarly, 89.3% of sheep and 100% of moufl on rams had Y chromosome haplotype H6 while seven, fi ve, three and two rams had haplotypes H8, H18, H7 and H5, respectively. Our study shows that present sheep populations of Eastern Adriatic, with exception of several rams that are indicating remote origin, are homogenous and belong to the lineages characteristic for European sheep populations.
Equine melanomas occur most commonly in grey horses at age 5 years or more. Generally, benign and... more Equine melanomas occur most commonly in grey horses at age 5 years or more. Generally, benign and malignant melanomas are distinguished by microscopy, but a more distinct classification would be helpful. The objectives of this study were to gain further evidence concerning the occurrence of melanotic tumours, and to evaluate the impact of heredity on melanoma development. A clinical study was conducted on a defined population of 296 grey horses of Lipizzaner breed. Individuals were classified according to their stage of disease using a 0-5 scale. Heritability was estimated on a sample of 296 grey horses with pedigrees traced back as far as 32 generations. Of the 296 horses, dermal melanomas were present in 148 horses (50%), 68 of which were more than age 15 years; 51 of these were melanoma-bearing. In 75.6% of cases, melanotic tumours were detected underneath the tail. Although melanoma-bearing grey horses were encountered up to stage 4, none of the affected individuals suffered any severe clinical effect or was handicapped in performance. Statistical analysis revealed highly significant effects of stud and age (P < 0.0001), explaining 28% of the total variability. In contrast to melanomas in solid-coloured horses characterised by early metastases, melanomas in grey horses showed less malignancy. Affected individuals often had encapsulated nodules or structures similar to human blue nevi. Grey horse-specific genetic factors inhibiting metastatic processes may be responsible for this phenomenon. Although the obtained heritability estimate of 0.36 with a standard error of 0.11 indicates a strong genetic impact on the development of melanoma in ageing grey horses, a possible influence of the genes with large effects was also suggested. Therefore, further analysis is required of melanoma development in the ageing grey horse.
It is well known that cheese yield and quality are affected by animal genetics, milk quality (che... more It is well known that cheese yield and quality are affected by animal genetics, milk quality (chemical, physical, and microbiological), production technology, and the type of rennet and dairy cultures used in production. Major differences in the same type of cheese (i.e., hard cheese) are caused by the rennet and dairy cultures, which affect the ripening process. This review aims to explore current technological advancements in animal genetics, methods for the isolation and production of rennet and dairy cultures, along with possible applications of microencapsulation in rennet and dairy culture production, as well as the challenge posed to current dairy technologies by the preservation of biodiversity. Based on the reviewed scientific literature, it can be concluded that innovative approaches and the described techniques can significantly improve cheese production.
Proceedings of 12th World Congress on Genetics Applied to Livestock Production (WCGALP)
Livestock Guardian Dogs (LGDs, Canis familiaris) are large dogs adapted to harsh mountain environ... more Livestock Guardian Dogs (LGDs, Canis familiaris) are large dogs adapted to harsh mountain environments and selected to protect grazing livestock from predators in free-ranging grazing systems. Signals of positive selection in a meta-population represented by nine LGD breeds (116 individuals) were analysed based on information from 104,618 SNP genotypes using the extreme Runs of Homozygosity islands (eROHi) and the integrated Haplotype Score (iHS) approach. We found that 28 genomic regions (11 identified using the eROHi approach and 17 using the iHS approach) had highly significant signals for positive selection. Some of the most pronounced selection signals are associated with genes such as LRIG3, TRIP11, CNTNAP2 and EFNA5. The presented results will contribute to the understanding of the synergistic behavioural interaction of LGDs with humans and grazing animals, especially sheep and goats, and the mammals' adaptation to harsh mountain conditions, as well as to their breeding management.
Uploads
Papers by Ino Curik