J. Anim Sci. 2007. 85:592-597. doi:10.2527/jas.2006-308
© 2007 American Society of Animal Science
Association analyses between the prion protein locus and reproductive and lamb weight traits in Ripollesa sheep1
J. Casellas*,
,
G. Caja*,,2,
R. Bach
,
O. Francino
, and
J. Piedrafita*,
* Grup de Recerca en Remugants,
and
Servei Veterinari de Genètica Molecular,
and
Departament de Ciència Animal i dels Aliments, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain; and
and
Associació Nacional de Criadors d’Ovins de Raça Ripollesa, 17121 Monells, Girona, Spain
 |
Abstract
|
|---|
The objective of this study was to analyze the association between the haplotypes of the prion protein (PrP) locus and several reproductive and lamb weight traits in Ripollesa sheep. Prion protein genotypes were available for a total of 310 sheep (7 rams, 114 ewes, and 189 lambs), all of them belonging to the purebred Ripollesa flock of the Universitat Autònoma of Barcelona, for which all sheep had a known pedigree. In addition, the genotype of 24 historical descendants of the previously genotyped adult individuals was reconstructed, provided that both parents were homozygous for PrP haplotypes. Only 3 haplotypes (ARR, ARQ, and ARH) were observed in the PrP locus of the sheep sampled. Reproductive traits included conception rate and litter size, whereas birth BW and 90-d BW were the lamb weight traits studied. The additive effect of PrP haplotypes was analyzed through Bayesian animal threshold and linear models, for reproduction and weight traits, respectively. Ewe reproductive data belonged to 89 ewes that gave 492 conception rate records and 440 litter size records. Analyses of BW at birth and at 90 d of age were made on 323 and 164 lamb records, respectively. No associations between PrP haplotypes and conception rate and BW traits were observed. For litter size, the effect of the ARH haplotype was greater than that of the ARQ haplotype. Differences between ARH and ARR haplotypes also suggested an advantage for the ARH. As a whole, our results indicated that the selection favorable to increase litter size in Ripollesa ewes may also increase the ARH haplotype frequency, which contradicts the recommendations of the current European Union legislation aiming to increase the genetic resistance to scrapie. As a consequence, scrapie genotyping needs to be included as a new selection criterion in the breed.
Key Words: birth weight • conception rate • litter size • prion protein • Ripollesa breed • sheep
|
INTRODUCTION
|
|---|
Scrapie is a fatal and transmissible prion disease of sheep and goat (Prusiner, 1998
). It is well known that ovine susceptibility is associated with polymorphisms in the prion protein (PrP) gene, located on chromosome 13 of sheep (Bossers et al., 1996). As Belt et al. (1995) indicated, up to 5 scrapie haplotypes have been described according to amino acids transcribed at codons 136, 154, and 171 (ARR, ARQ, ARH, AHQ, and VRQ). The greatest and lowest susceptibilities to scrapie have been related to VRQ and ARR haplotypes respectively (Jeffrey et al., 2001), although the ARQ haplotype also shows a high susceptibility (Bossers et al., 1996).
Breeding for scrapie resistance has been recently considered an attractive strategy to control this disease (Arnold et al., 2002). Indeed, selection for the ARR haplotype has been established by the European Commission (Commission Decision, 2003/100/CE) and the US government (Department of Agriculture, 2001), to increase the scrapie resistance of sheep breeds. Notwithstanding, breeding for PrP genotype could apply a high selection pressure on chromosome 13, but little is known about potential correlated effects on production or reproduction traits. Some studies have suggested an association among PrP haplotypes and litter size (Brandsma et al., 2004) and lamb BW at different ages (Buitkamp et al., 2003; Brandsma et al., 2004), although contradictory results have been reported (Roden et al., 2001; Ponz et al., 2004). In this regard, disagreement exists about the suitability of fixation or removal of the PrP, due to the possible correlated influences on production or reproductive traits.
The aim of this research was to analyze the association between PrP haplotypes and reproduction and lamb BW traits to investigate the effects of the PrP locus and to update the selection aims within the Ripollesa meat sheep breed.
|
MATERIALS AND METHODS
|
|---|
Animal data utilized in this experiment were obtained from a sheep flock and were collected according to the standards for the humane care and use of animals by the Ethical Committee of Animal and Human Experimentation of the Universitat Autonoma de Barcelona.
Genotyped Animals
The Ripollesa is a Spanish indigenous sheep breed of medium frame (ewes, 50 to 65 kg of BW; rams 75 to 90 kg of BW), exploited under semiintensive Mediterranean conditions for the production of light Pascual-type lambs (22 to 24 kg of BW at slaughter) in Catalonia, Spain. Since 1986, the Universitat Autònoma de Barcelona (Bellaterra, Spain) has kept a purebred Ripollesa flock with a size of approximately 100 ewes. After 1993, and with the exception of 2 rams bought in 1997, the flock was closed, with all subsequent ewe and ram replacements generated within the flock, after following a phenotypic selection program for litter size.
In 2004, all adult individuals (7 rams and 114 ewes) and 189 lambs born during the fall-lambing season of 2004 and 2005 were genotyped for the PrP locus. Blood samples from each animal were collected by jugular venipuncture into EDTA-coated Vacutainer tubes (Becton Dickinson Co., Franklin Lakes, NJ). The polymorphisms at codons 136, 154, and 171 of the PrP gene were determined by SNaPshot Multiplex analysis (Applied Biosystems, Foster City, CA) according to the methodology of Álvarez et al. (2006a). Analyses were performed in the laboratories of the Servei Veterinari de Genètica Molecular (Universitat Autònoma de Barcelona, Bellaterra, Spain) and Laboratorio Central Veterinario (Algete, Madrid, Spain). In addition, the genotypes of 24 historical animals were reconstructed from the knowledge that both parents were homozygous for the PrP locus.
Traits Analyzed and Operational Models
Two reproduction traits were considered (i.e., conception rate and litter size). Conception rate was scored with 1 if the ewe lambed during a given lambing season or 0 if it failed to lamb. Litter size was defined as 1 or 2, for single or multiple births, respectively. Triplet births were grouped with twin births due to their low incidence (2.7%). The model of analysis for both traits included the infinitesimal genetic effect of each animal (ai), as well as several nongenetic sources of variation: random permanent environment characterized by the ewe (pi); ewe age at lambing (EAj) with 5 categories following in part the ones stated by Matos et al. (1997) and Altarriba et al. (1998; i.e., 
2, 3, 4, 5, and 
6 yr); and year of lambing (YLk) with 7 levels (<2000<2000<2001<2002<2003<2004, and 2005). The genetic effect of each PrP haplotype was also included as a covariate with the number of copies of each haplotype, 0, 1, or 2 (ARRi, ARQi, and ARHi). Note that we assumed an additive model following in part the standard model of Falconer and Mackay (1996). Within the threshold methodology framework, the operational model was
 |
where µ was the population mean; ßARR, ßARQ, and ßARH were the corresponding regression coefficients for each PrP haplotype; and µijkl was the value of the underlying variable related to the observed categorical trait (conception rate or litter size) by thresholds, as defined by Wright (1934). Birth BW and 90-d BW were also analyzed although under a linear animal model. The operational models for these traits included the sex of the lamb (SXm) and the type of suckling (single or multiple; TSn) in addition to the effects previously defined for reproductive traits. The birth BW was also considered as a covariate (ßBBWBBWi within the 90-d BW model. The operational model was
where yijklmno was the birth BW (
= 0) or 90-d BW ( = 1) phenotypic record.
Bayesian Statistical Analysis
All traits were analyzed under univariate animal models solved through a Bayesian approach. More specifically, birth BW and 90-d BW were assumed to follow a multivariate normal distribution y 
N(W
, I
2e), where y was the vector of phenotypic values, W was an incidence matrix relating systematic effects (ß), permanent environmental effects (p), and additive genetic effects (a), represented by the vector ' = [ß' p' a'], I was an identity matrix, and 2e was the residual variance. Prior distributions for the infinitesimal genetic effects (a) and permanent environmental effects (p) were described as multivariate normal densities:
where A was the numerator relationship matrix, and 2a and 2p represented the additive genetic and permanent environmental variances, respectively. Proper uniform priors between adequate values were defined for the ß systematic effects and variance components.
On the other hand, conception rate and litter size were analyzed using the threshold procedure described by Sorensen et al. (1995). This methodology related the discrete phenotypic observation with an underlying continuous variable, also called liability (u). Liability was assumed multivariate normal, u N(W, I2e), and the remaining parameters of the model were fitted as for the linear traits described previously, with the exception of 2e being fixed at 1, and the threshold between the 2 phenotypic categories being fixed at zero in the unobservable scale (Sorensen et al., 1995).
Marginal posterior distributions of all parameters were obtained using the Gibbs sampler (Gelfand and Smith, 1990). Note that Gibbs iterations for threshold models required an extra step known as data augmentation (Tanner and Wong, 1987) that allowed the inclusion of liability as an unknown parameter in the model (Sorensen et al., 1995). In the present work, the Gibbs sampler ran with a single chain of 100,000 points, and the first 10,000 were discarded as burn-in, previously tested by the Raftery and Lewis (1996) methodology.
Our Bayesian approach considered the marginal posterior distribution of the differences between each pair of additive effects (ßARR – ßARQ, ßARR – ßARH, and ßARQ – ßARH). In this case, the posterior probability above or below zero was the probability of a difference bigger or smaller than zero, respectively, given the data. Substantial evidence of dissimilarity between the additive effect of 2 haplotypes was assumed when the null difference was not included within the 95% greatest posterior density interval. Although prior distributions have an important role in the inference of the posterior distribution, the prior information for the systematic effects of our models was vague with null or very low influence in the posterior distributions. Moreover, the inference was focused in the systematic effects characterized by the PrP haplotypes, considering as nuisance parameters the rest of unknowns of the model. For this reason, the posterior inference was robust.
|
RESULTS
|
|---|
The Ripollesa flock studied showed only 3 haplotypes of the PrP locus, ARR, ARQ, and ARH (Table 1) with frequencies of 0.34, 0.54, and 0.13, respectively, within the group of adult individuals genotyped in 2004. These frequencies were similar to the ones obtained in the whole data of the Ripollesa breed in Spain, which also showed a minimal incidence of AHQ (0.02) and VRQ (0.03) haplotypes (Table 1).
The number of animals and records used to study reproductive and weight traits in the Ripollesa breed are shown in Table 2. The phenotypic records analyzed showed an average conception rate of 0.89, with 1.61 lambs born per birth. Lambs were born with an average birth weight of 3.22 kg, reaching 19.89 kg at an age of 90 d (Table 2).
Marginal posterior densities of the differences between the additive effects of PrP haplotypes are shown in Table 3. The null difference falls within the greater posterior density region at 95% (HPD95) for all haplotype differences corresponding to conception rate, birth weight, and 90-d BW traits (Table 3). Our results showed that the ARH haplotype increased litter size in contrast with the ARQ haplotype. The mode of the difference between both haplotypes was 0.41 units of residual variance favorable to ARH, its HPD95 (0.11 to 0.75) being far from the null difference (Table 3). On the other hand, the marginal posterior density for the difference between ARH and ARR haplotypes within litter size also suggested an advantage for the ARH haplotype, although the HPD95 ranged between –0.02 and 0.72. Furthermore, the difference between ARR and ARQ haplotypes was almost null for litter size (Table 3). Finally, the predicted incidences of multiple births in our flock after substituting the current genotype of each ewe by ARR/ARR, ARQ/ARQ, or ARH/ARQ (see Appendix) were 0.59, 0.58, and 0.62, respectively (Figure 1).
View this table:
[in this window]
[in a new window]
|
Table 3. Mode and 95% greatest posterior density region bounds (HPD95) of the differences between additive effects of each pair of haplotypes
|
|
View larger version (21K):
[in this window]
[in a new window]
|
Figure 1. Predicted incidence of multiple births in the flock after substituting the current genotype of each ewe by ARR/ARR, ARQ/ARQ, and ARH/ARQ. The vertical dashed line identifies the observed incidence of multiple births in the data set.
|
|
|
DISCUSSION
|
|---|
Only 3 different haplotypes were found in our Ripollesa breed flock, ARR, ARQ, and ARH, whereas the haplotypes VRQ and AHQ described by Belt et al. (1995) were not detected. Both these haplotypes have been identified in the Ripollesa breed population in Spain, although their joint frequency was low, close to 0.05 (Table 1
). The reduced frequency of the VRQ haplotype might reduce the potential susceptibility of Ripollesa breed to scrapie as indicated by Hunter et al. (1994). Unfortunately, the ARQ haplotype also contributes an important degree of scrapie susceptibility (Bossers et al., 1996), and this haploype was the most abundant in the Ripollesa breed. The frequency of the ARR haplotype corresponding to the adult individuals of the genotyped flock has reached a value of 0.34, greater than the values reported in East Friesian dairy sheep (De Vries et al., 2005) and the Spanish Rasa Aragonesa (Ponz et al., 2004), close to the value described in Texel (Brandsma et al., 2004), and lower than the value observed in several Bavarian sheep breeds (Buitkamp et al., 2003). The moderate ARR haplotype frequency reported might allow the breeding of Ripollesa sheep breed toward scrapie resistance following the European Union legislation (Commission Decision, 2003/100/CE), although accurate selection strategies should be implemented to avoid excessive inbreeding (Sonesson et al., 2003), bottleneck effects, and drop of genetic progress on productive traits (Álvarez et al., 2006b). In this sense, a minimum ARR frequency of 0.25 has been suggested to establish a selection program for resistance to scrapie (Álvarez et al., 2006a).
The effect of the different PrP haplotypes on conception rate and litter size was tested. The differences between additive effects on conception rate were null. This fact could be related to the low heritabilities reported for this trait (Fogarty, 1995). On the other hand, our analysis reported notable differences between haplotypes for litter size. The ARH haplotype showed a positive effect in contrast to the ARQ haplotype, and also seemed advantageous in relation to the ARR haplotype, showing that selection favorable to the ARH haplotype might increase litter size in our flock. Note that the ARH haplotype is usually associated with a substantial degree of genetic resistance to scrapie (Woolhouse et al., 2001). Predicted incidence of multiple births, calculated as indicated in the appendix, might be greater for ARH heterozygous ewes (0.62) than for ARR (0.59) and ARQ homozygous ewes (0.58), suggesting a genetic influence linked to PrP haplotype. The predicted incidence of multiple births was not estimated for ARH homozygous ewes because we had no ARH homozygous ewes genotyped in our flock, and then, the reported advantage of the ARH haplotype may also include dominance effects (Figure 1). The magnitude of the litter size increase was in accordance with the results reported by Alexander et al. (2005) in the Suffolk breed, although these authors compared the ARR haplotype with all the remaining ones.
Several authors have described influences of the PrP locus on ewe litter size although different haplotypes have been reported as favorable in the different breeds analyzed. Our results are in accordance with Ponz et al. (2004) who found a slight genetic advantage for ewes with the ARH haplotype in the Rasa Aragonesa breed, which is geographically close to the Ripollesa breed. On the other hand, Brandsma et al. (2004) suggested a slight advantage for the ARR haplotype in the Texel breed, although they analyzed ARH and ARQ haplotypes jointly, within the same category. Alexander et al. (2005) reported the opposite effect in the Suffolk breed, and no significant influences in Columbia, Rambouillet, and Hampshire sheep. Similarly, De Vries et al. (2005) also reported no significant effects of PrP haplotypes on ewe prolificacy. As a whole, although pleiotropy of the PrP gene has been suggested as a putative mechanism to influence litter size, the discrepancies between previously published results (Brandsma et al., 2004; Ponz et al., 2004; Alexander et al., 2005) suggest that this hypothesis may be wrong. It seems more reasonable to assume that genetic influences might proceed from a gene closely linked to the PrP locus, within the ovine chromosome 13.
The PrP haplotypes did not modify birth BW and 90-d BW in the Ripollesa breed (Table 3). Although litter size greatly influences lamb birth BW, associations between PrP haplotypes and this trait have not been detected, in contrast with the results corresponding to ewe litter size. Besides, litter size was included as a systematic effect within the birth BW model. The absence of association with the 90-d BW of Ripollesa lambs is in accordance with the results reported for lamb BW at different ages in Suffolk (Roden et al., 2001). On the contrary, Buitkamp et al. (2003) reported a positive effect of the ARR haplotype on ADG of Merino lambs, although similar differences could not be detected in Blackface and Suffolk lambs. Significant influences also were reported by Brandsma et al. (2004) who observed a reduction of the BW at 135 d in the Texel breed associated with the ARR haplotype.
Recent legislation of the European Union recommended selection for ARR haplotype as a measure to increase the genetic resistance of sheep breeds to scrapie. Unfortunately, our results show that the fixation of the ARR haplotype may reduce the reproductive performance of the Ripollesa ewe, due to the loss of the ARH haplotype. Although future studies are necessary to determine if our results could be extrapolated to the entire Ripollesa population, they contribute vital information for the breeding program of this breed. They also show that marker-assisted selection using the PrP locus would be an additional tool for ewe prolificacy improvement if current legislation allowed selection favorable to the ARH haplotype.
|
IMPLICATIONS
|
|---|
Associations between prion protein haplotypes and several reproductive and weight traits were tested in a flock of Ripollesa breed, and substantial differences between haplotypes were not observed for conception rate and weight traits. However, the ARH haplotype clearly increased litter size in contrast with the ARQ haplotype, and the ARH haplotype also suggested a moderate advantage over the ARR haplotype. Differences between ARR and ARQ haplotypes were close to zero. The current results suggested that a breeding program favorable to the ARH haplotype might increase ewe reproductive performance in the Ripollesa breed. On the contrary, the selection favorable to the ARR haplotype to increase the genetic resistance to scrapie could reduce ewe prolificacy, although future studies are required to determine if our results may be extrapolated to the entire Ripollesa population and to characterize the genetic architecture of the prion protein locus in the ovine species.
|
APPENDIX
|
|---|
Under the threshold models methodology (Wright, 1934; Sorensen et al., 1995), solutions obtained for the genetic effect of each PrP haplotype predict the net increase (or decrease) of liability (u) due to the presence of each haplotype within the genotype of an animal. Nevertheless, these solutions cannot be easily transformed to an observable scale, where their magnitude depends on the remaining effects that influence each animal. Taking as an example the ewe prolificacy trait and given the Gaussian distribution assumed for its liability u N(W, I2e), the probabilities to achieve a multiple birth under the influence of the jth (PMBj) or kth (PMBk) scalar elements in the vector , both corresponding to mutually exclusive categories of the same systematic effect S, are
where w'–S is the incidence vector of –S, both with all elements belonging to the categories of S excluded; and j and k are the jth and kth scalar elements in the vector , respectively. Note that the threshold for liability has been fitted to zero according to Sorensen et al. (1995). The advantage of j with respect to k in the observable scale can be characterized as the difference PMBj – PMBk, although it is highly influenced by the incidence assumed for the remaining systematic, permanent, and additive genetic effects (w'–S–S). In this sense, PMBj – PMBk reaches its greatest value when w'–S–S + k = –(j –k)/2, whereas it becomes zero when w'–S–S tends to infinity. As a whole, this approach may only be useful if we can define a representative incidence of the remaining systematic effects.
To compare the effect of j and k in the observable scale and bypassing the subjectivity of an assumed and unique w'–S, we can predict twice the liability value for each record i of our data set, taking its own w'i,–S and assuming the influence of j (ui,j) and k (ui,k) separately. It can be estimated as:
Thus, we obtain 2 liability sets, one under the influence of j and the other receiving the k effect. Considering that the threshold for liability has been fitted to zero, we can calculate the predicted incidence of multiple births as the frequency above zero liabilities for each data set. The difference between both data sets predicts the advantage or disadvantage of j with respect to k for the characteristic incidence of systematic, permanent, and additive genetic effects of our population.
|
Footnotes
|
|---|
1 Research supported by a contract with the Departament d’Agricultura, Ramaderia i Pesca de la Generalitat de Catalunya (Spain) and a Universitat Autònoma de Barcelona (Bellaterra, Spain) fellowship granted to J. Casellas. The authors appreciate the assistance of R. Costa and the crew of the S1GCE (Servei de Granges i Camps Experimentals de la UAB, Bellaterra, Spain) for feeding and taking care of the animals, and to Nic Aldam for the English revision of the manuscript. 
2 Corresponding author: gerardo.caja{at}uab.es
Received for publication May 12, 2006.
Accepted for publication October 17, 2006.
|
LITERATURE CITED
|
|---|
Alexander, B. M., R. H. Stobart, W. C. Russell, K. I. O’Rourke, G. S. Lewis, J. R. Logan, J. V. Duncan, and G. E. Moss. 2005. The incidence of genotypes at codon 171 of the prion protein gene (PRNP) in five breeds of sheep and productive traits of ewes associated with those genotypes. J. Anim. Sci. 83:455–459.[Abstract/Free Full Text]
Altarriba, J., L. Varona, L. A. García-Cortés, and C. Moreno. 1998. Bayesian inference of variance components for litter size in Rasa Aragonesa sheep. J. Anim. Sci. 76:23–28.[Abstract/Free Full Text]
Álvarez, L., J. J. Arranz, and F. San Primitivo. 2006a. Identification of a new leucine haplotype (ALQ) at codon 154 in the ovine prion protein gene in Spanish sheep. J. Anim. Sci. 84:259–265.[Abstract/Free Full Text]
Álvarez, L., B. Gutiérrez-Gil, F. San Primitivo, L. F. De la Fuente, and J. J. Arranz. 2006b. Influence of prion protein genotypes on milk production traits in Spanish Churra sheep. J. Dairy Sci. 89:1784–1791.[Abstract/Free Full Text]
Arnold, M., C. Meek, C. R. Webb, and L. J. Hoinville. 2002. Assessing the efficacy of a ram-genotyping programme to reduce susceptibility to scrapie in Great Britain. Prev. Vet. Med. 56:227–249.[CrossRef][Medline]
Belt, P. B., I. H. Muileman, B. E. Schreuder, J. Bos-de Ruijter, A. L. Gielkens, and M. A. Smits. 1995. Identification of five allelic variants of the sheep PrP gene and their association with natural scrapie. J. Gen. Virol. 76:509–517.[Abstract/Free Full Text]
Bossers, A., B. E. C. Schreuder, I. H. Muileman, P. Belt, and M. A. Smits. 1996. PrP genotype contributes to determining survival times of sheep with natural scrapie. J. Gen. Virol. 77:2669–2673.[Abstract/Free Full Text]
Brandsma, J. H., L. L. G. Janss, and A. V. Visscher. 2004. Association between PrP genotypes and litter size and 135 days weight in Texel sheep. Livest. Prod. Sci. 85:59–64.[CrossRef]
Buitkamp, J., R. Emmerling, J. Semmer, C. Mendel, M. Wagenpfeil, A. Steiner, and K. U. Götz. 2003. Association of PRnP-Alleles with growth performance and carcass traits in Bavarian sheep breeds. Proc. Int. Workshop on Major Genes and QTL in Sheep and Goats, Tolouse, France, Communication n° 2–29.
Commission Decision (2003/100/EC) of February 2003. Laying Down Minimum Requirements for the Establishment of Breeding Programmes for Resistance to Transmissible Spongiform Encephalopathies in Sheep (Text with EEA relevance) (notified under document number C (2003) 498). Official J. L 041, 14/02/2003:41–45.
Department of Agriculture. 2001. Scrapie in sheep and goats; Interstate movement restrictions and indemnity program. Fed. Regist. 66:43963–44003.
De Vries, F., H. Hamann, C. Drögemüller, M. Ganter, and O. Distl. 2005. Analysis of associations between the prion protein genotypes and production traits in East Friesian Milk sheep. J. Dairy Sci. 88:392–398.[Abstract/Free Full Text]
Falconer, D. S., and T. F. C. Mackay. 1996. Introduction to Quantitative Genetics. Longman, NY.
Fogarty, N. M. 1995. Genetic parameters for live weight, fat and muscle measurements, wool production and reproduction in sheep: A review. Anim. Breed. Abstr. 63:101–143.
Gelfand, A., and A. F. M. Smith. 1990. Sampling based approaches to calculating marginal densities. J. Am. Stat. Assoc. 85:398–409.[CrossRef]
Hunter, N., W. Goldmann, G. Smith, and J. Hope. 1994. The association of a codon-136 PrP gene variant with the occurrence of natural scrapie. Arch. Virol. 137:171–177.[CrossRef][Medline]
Jeffrey, M., S. Ryder, S. Martin, S. A. Hawkins, L. Terry, C. Berthelin-Baker, and S. J. Bellworthy. 2001. Oral inoculation of sheep with the agent of bovine spongiform encephalopathy (BSE). 1. Onset and distribution of disease-specific PrP accumulation in brain and viscera. J. Comp. Pathol. 124:280–289.[CrossRef][Medline]
Matos, C. A. P., D. L. Thomas, D. Gianola, R. J. Tempelman, and L. D. Young. 1997. Genetic analysis of discrete reproductive traits in sheep using linear and nonlinear models: I. Estimation of genetic parameters. J. Anim. Sci. 75:76–87.[Abstract/Free Full Text]
Ponz, R., T. Tejedor, A. Laviña, and M. V. Arruga. 2004. Estudio de la asociación de los alelos del gen PrP con el carácter prolificidad en la raza Rasa Aragonesa. Pages 368–370 in Proc. XXIX Jornadas Científicas SEOC, Lleida, Spain. Sociedad Española de Ovinotecnia y Caprinotecnia, Lleida, Spain.
Prusiner, S. B. 1998. Prions. Proc. Natl. Acad. Sci. USA 95:13363–13383.[Abstract/Free Full Text]
Raftery, A. E., and S. M. Lewis. 1996. Implementing MCMC. Pages 115–130 in Markov Chain Monte Carlo in practice. W. R. Gilks, S. Richardson, and D. J. Spiegelhalter, ed. Chapman & Hall, London, UK.
Roden, J. A., W. Haresign, and J. M. L. Anderson. 2001. Analysis of PrP genotype in relation to performance traits in Suffolk sheep. Page 45 in Proc. Br. Soc. Anim. Sci., London, UK. Br. Soc. Anim. Sci., London, UK.
Sonesson, A. K., L. L. G. Janss, and T. H. E. Meuwissen. 2003. Selection against genetic defects in conservation schemes while controlling inbreeding. Genet. Sel. Evol. 35:353–368.[Medline]
Sorensen, D. A., S. Andersen, D. Gianola, and I. Korsgaard. 1995. Bayesian inference in threshold models using Gibbs sampling. Genet. Sel. Evol. 27:229–249.[CrossRef]
Tanner, M. A., and W. H. Wong. 1987. The calculation of posterior distributions by data augmentation. J. Am. Stat. Assoc. 82:528–550.[CrossRef]
Woolhouse, M. E. J., P. Coen, L. Matthews, J. D. Foster, J. M. Elsen, R. M. Lewis, D. T. Haydon, and N. Hunter. 2001. A centuries-long epidemic of scrapie in British sheep? Trends Microbiol. 9:67–70.[CrossRef][Medline]
Wright, S. 1934. An analysis of variability in number of digits in an inbred strain of Guinea pigs. Genetics 19:506–536.[Free Full Text]
Top
Abstract
INTRODUCTION
