ABSTRACT We aimed to reduce the dimensionality of quantitative traits in pigs applying factor analysis. Quantitative variables were collected, and the factor analysis extracted five factors with biological meaning related to performance, carcass quality, carcass yield, meat quality, and initial pH. These factors were posteriorly used as dependent variable to evaluate the effects of genetic groups (Piau, Duroc, and Pietrain crossbreds), nutritional plans (low, medium and high lysine levels) and sex. An interaction effect between genetic group and sex was observed in performance, in which Duroc crossbred showed the greatest scores. The Pietrain genetic group showed greater values of carcass quality compared with Duroc and Piau, while Piau crossbred pigs had greater values for meat quality compared with Pietrain and Duroc. A greater carcass yield was observed in Pietrain crossbred compared with Duroc pigs. Pigs fed under low lysine level had the lowest performance values. Carcass yield was affected by nutritional plans, in which the medium lysine level showed the greatest values. Meat quality was improved by the nutritional plan with high lysine level compared with the low lysine level. Initial pH showed increased means when using nutritional plans with low and medium lysine levels in diet composition. With regard to sex effects, a greater carcass quality was found for gilts compared with barrows. The reduction of the data dimensionality allowed the joint evaluation of genetic group, nutritional plan, and sex based on new latent variables that represented the original dataset under easier framework based on factors’ biological interpretation.

ABSTRACT We aimed to estimate genetic parameters for growth, reproductive, and carcass traits in Tabapuã cattle. Phenotypic data were collected between 1990 and 2019 in 1,218 farms, and the pedigree file had 340,868 animals. The traits evaluated were body weight at 120 (W120), 210 (W210), 365 (W365), and 550 (W550) days of age; age at first calving (AFC), scrotal circumference at 365 days of age (SC365), ribeye area (REA), backfat thickness (BF), and rump fat thickness (RF). The (co)variance components were estimated using the restricted maximum likelihood method, considering single and two-traits animal models. For all traits, the models considered fixed, direct additive genetic, and residual random effects. In addition, for W120 and W210, the maternal additive genetic and maternal permanent environmental effects were also included. Heritabilities for W120, W210, W365, W550, SC365, REA, BF, and RF were of moderate magnitude (0.15, 0.16, 0.23, 0.19, 0.22, 0.36, 0.31, and 0.27, respectively). Low heritability was observed for AFC (0.07). The genetic correlations between growth traits were higher than 0.90, while AFC and SC365 presented negative moderate correlation (−0.66). The REA showed low genetic correlations with BF (0.07) and RF (0.07), whereas BF and RF were highly correlated (0.77). Considering the heritability estimates, selection for AFC would result in limited genetic gain, while for the other traits, it would be satisfactory. Based on the high genetic correlations between growth traits, selection of Tabapuã animals can be performed at younger ages. Additionally, animals can be indirectly selected for AFC through SC365, and only one fat thickness trait may be used in the selection process considering the high genetic correlation and similar heritability values for BF and RF.

ABSTRACT We aimed to estimate litter size and litter uniformity genetic parameters and genetic trends of Landrace pigs at birth and at three weeks by using multitrait analyses for 2,787 litters. The following litter traits were evaluated: number of piglets born alive (NBA), within-litter weight mean at birth (MBW), within-litter weight standard deviation at birth (SDB), within-litter weight coefficient of variation at birth (CVB), number of piglets at three weeks (NT), within-litter weight mean at three weeks (MT), within-litter weight standard deviation at three weeks (SDT), and within-litter weight coefficient of variation at three weeks (CVT). Heritability estimates for NBA, MBW, SDB, and CVB were 0.09±0.04, 0.31±0.08, 0.01±0.04, and 0.07±0.05, respectively, greater than those obtained at three weeks (0.06±0.04, 0.10±0.06, 0.01±0.04, and 0.02±0.04 for NT, MT, SDT, and CVT, respectively). The genetic correlations between NBA and MBW and between MBW and CVB (−0.73±0.20 and −0.93±0.21, respectively) were of moderate to high magnitudes, as well as the genetic correlations between CVT and SDT (0.85±0.39). Genetic correlations between MBW and MT, SDB and SDT, CVB and CVT, and NBA and NT were 0.73±0.16, 0.69±0.54, 0.36±0.80, and 0.95±0.06, respectively. The genetic trends were linear for NBA and CVB and quadratic for MBW and SDB, whereas for all traits at three weeks, they were close to zero. Within-litter weight coefficient of variation (CV) may be the most appropriate variation measure for application in breeding programs, especially at birth, due to its greater heritability estimate and high and negative genetic correlation with MBW. The genetic trends show that NT does not follow the increase in NBA, emphasizing the need to review the breeding goals.

Abstract We aimed to evaluate the effects of paternal genotypes (Piau, Duroc-based, and Pietrain) and nutritional plans with different digestible lysine contents (Low, Medium, and High) on performance, carcass, and meat quality traits of pigs. Pietrain and Duroc crossbred barrows and gilts were similar in performance and carcass traits and had the greatest values when compared to Piau crossbred pigs. Regarding meat quality, drip loss was greater in pork from Pietrain crossbred barrows and gilts compared to Duroc and Piau crossbred pigs. The High nutritional plan had the greatest daily digestible lysine intake value, followed by the Medium and Low nutritional plans. Most of the performance, carcass, and meat quality traits that were evaluated in barrows and gilts were not affected by the nutritional plans. In general, Duroc and Pietrain crossbred pigs had a greater performance and carcass yield when compared to Piau crossbred pigs.

ABSTRACT: We evaluated the inclusion of information on genetic relationship into the analysis of crude protein requirement in diets for pigs of Brazilian Piau breed, using Bayesian inference. The animals were assigned to treatments in a completely randomized design in factorial scheme 4 × 2 (crude protein levels × sex) with 12 repetitions per treatment. The evaluations were carried out in the initial, growing and finishing phases, and after slaughter. The traits evaluated were feed conversion (FC), backfat thickness (BF), daily weight gain (DWG), daily feed intake (DFI) and some carcass cuts. Three models were considered to evaluate the inclusion of information on genetic relationship into the analysis: Model I, a simple linear model; Model II, the same effects of Model I with addition of the independent random effect of animal; and Model III, the same effects of Model II, but including the genetic relationship between the animals. Model III presented the best fit and was considered for later inferences. Crude protein (CP) levels did not significantly influence any of the evaluated traits. The effect of sex was significant only for the growing phase, while its interaction with protein levels presented an opposite result for all evaluated traits. Additionally, CP levels of 10.2 %, 9.6 % and 9.0 % can be used in diets for pigs of Brazilian Piau breed in the initial, growing and finishing phases, respectively.

Resumo: O objetivo deste trabalho foi avaliar o uso da regressão quantílica regularizada (RQR) para predição do mérito genético em suínos quanto a características assimétricas de carcaça, em comparação ao método do lasso bayesiano (Blasso). Foram utilizados dados genéticos das características rendimento de carcaça, espessura de bacon e espessura de toucinho de uma população F2 composta por 345 indivíduos, obtida pelo cruzamento de animais da raça Piau com os de raça comercial. A RQR foi avaliada tendo-se considerado diferentes quantis (τ = 0,05 a 0,95). O modelo de RQR utilizado para estimar o mérito genético apresentou acurácias maiores ou iguais àquelas obtidas pelo Blasso, para todas as características estudadas. Houve incremento de 6,7 e 20,0% na acurácia, quando se consideraram os quantis 0,15 e 0,45 na avaliação do rendimento de carcaça e da espessura de bacon, respectivamente. Os resultados obtidos são indicativos de que a regressão quantílica regularizada apresenta maior acurácia que o método lasso Bayesiano para predição do mérito genético de variáveis assimétricas de carcaça.

Abstract: The objective of this work was to evaluate the use of regularized quantile regression (RQR) to predict the genetic merit of pigs for asymmetric carcass traits, compared with the Bayesian lasso (Blasso) method. The genetic data of the traits carcass yield, bacon thickness, and backfat thickness from a F2 population composed of 345 individuals, generated by crossing animals from the Piau breed with those of a commercial breed, were used. RQR was evaluated considering different quantiles (τ = 0.05 to 0.95). The RQR model used to estimate the genetic merit showed accuracies higher than or equal to those obtained by Blasso, for all studies traits. There was an increase of 6.7 and 20.0% in accuracy when the quantiles 0.15 and 0.45 were considered in the evaluation of carcass yield and bacon thickness, respectively. The obtained results are indicative that the regularized quantile regression presents higher accuracy than the Bayesian lasso method for the prediction of the genetic merit of pigs for asymmetric carcass variables.

ABSTRACT We aimed to evaluate the impact of using embryo transfer (ET) information on weaning weight estimated breeding values (EBV) and its accuracy. Data from Hereford and Braford cattle, raised under extensive conditions in Southern Brazil, were used. A model that included ET information in addition to maternal (genetic and permanent environmental) effects as a function of foster dams was compared to a model without ET information. Accuracy of both bulls and calves increased due to inclusion of ET records in 0.04 and 0.12 points, respectively. In general, the inclusion of ET records provided a greater amount of phenotypic variance and most accurate EBV for sires and progeny. The results obtained in this study encourage the use of ET phenotypic records in large-scale genetic evaluation programs, especially for bulls that have most of their progeny coming from ET. Most of the Brazilian genetic evaluation programs do not use phenotypic records of ET animals. Therefore, breeding values are predicted based only on parentage average, which implies in underestimated accuracies. Considering that ET has been widely used in Brazil and that such information improves genetic predictions, we suggest modifying the traditional adopted models by considering ET information in the Brazilian genetic evaluations.

Resumo: O objetivo deste trabalho foi avaliar os critérios para formação de grupos de contemporâneos (GCs) na avaliação genética de peso ao desmame em bovinos Nelore. Foram utilizados 713.474 registros de 3.066 rebanhos localizados no Centro-Oeste e no Norte do Brasil. Os dados foram provenientes do Serviço de Registro Genealógico das Raças Zebuínas da Associação Brasileira dos Criadores de Zebu. As estruturas de dados foram definidas com base no número de desvios-padrão (DPs) para remoção de outliers (±2,0, ±2,5, ±3,0 e ±3,5) e no número mínimo de animais por GC (3, 7 e 15). A avaliação genética foi realizada por meio de modelo animal via inferência bayesiana. As estruturas de dados com ±3,5 DPs e GC com mínimo de 15 animais apresentaram a maior variância genética aditiva (82,65±2,93), e aquelas com ±2,0 DPs e GC com mínimo de 3 animais, a menor (60,23±1,96). A formação correta de GCs resulta em arquivos de dados mais consistentes, o que permite obter estimativas mais confiáveis dos parâmetros genéticos. As melhores respostas à seleção são obtidas ao se adotar os seguintes critérios para remoção de outliers: 2,5, 3,0 e 3,5 desvios-padrão e mínimo de 15 animais por grupo de contemporâneos.

Abstract: The objective of this work was to evaluate the criteria for the formation of contemporary groups (CGs) in the genetic evaluation of body weight at weaning in Nellore cattle. A total of 713,474 records from 3,066 herds located in Midwestern and Northern Brazil were used. Data were obtained from the genealogical registry of zebu breeds of the Brazilian association of zebu breeders. Data structures were defined based on the number of standard deviations (SDs) for outlier removal (±2.0, ±2.5, ±3.0, and ±3.5) and on the minimal number of animals per CG (3, 7, and 15). Genetic evaluation was performed with an animal model using Bayesian inference. Data structures with ±3.5 SDs and CG with at least 15 animals presented the highest additive genetic variance (82.65±2.93), and those with ±2.0 SDs and CG with at least 3 animals showed the lowest one (60.23±1.96). The proper formation of CGs results in better-quality data archives, allowing to obtain more trustable estimates for the genetic parameters. Better selection responses are obtained when the following criteria are adopted for the removal of outliers: 2.5, 3.0, and 3.5 standard deviations and a minimum of 15 animals per contemporary group.

ABSTRACT: Genome association analyses have been successful in identifying quantitative trait loci (QTLs) for pig body weights measured at a single age. However, when considering the whole weight trajectories over time in the context of genome association analyses, it is important to look at the markers that affect growth curve parameters. The easiest way to consider them is via the two-step method, in which the growth curve parameters and marker effects are estimated separately, thereby resulting in a reduction of the statistical power and the precision of estimates. One efficient solution is to adopt nonlinear mixed models (NMM), which enables a joint modeling of the individual growth curves and marker effects. Our aim was to propose a genome association analysis for growth curves in pigs based on NMM as well as to compare it with the traditional two-step method. In addition, we also aimed to identify the nearest candidate genes related to significant SNP (single nucleotide polymorphism) markers. The NMM presented a higher number of significant SNPs for adult weight (A) and maturity rate (K), and provided a direct way to test SNP significance simultaneously for both the A and K parameters. Furthermore, all significant SNPs from the two-step method were also reported in the NMM analysis. The ontology of the three candidate genes (SH3BGRL2, MAPK14, and MYL9) derived from significant SNPs (simultaneously affecting A and K) allows us to make inferences with regards to their contribution to the pig growth process in the population studied.

RESUMO: A principal contribuição da seleção genômica (SG) é a identificação de animais geneticamente superiores para características de difícil mensuração e/ou avaliadas tardiamente nos animais, tal como o pH da carne suína. Na SG, uma vez que o número de marcadores é geralmente maior que o número de animais genotipados, e tais marcadores são altamente correlacionados (devido ao desequilíbrio de ligação), métodos estatísticos baseados na redução de dimensionalidade têm sido propostos. Dentre estes, destaca-se o Quadrados Mínimos Parciais (PLS) pela simplicidade e alta acurácia de predição. Porém, a determinação do número ótimo de componentes a ser utilizado no PLS ainda se caracteriza como um desafio para a aplicação do método. Assim, objetivou-se aplicar o PLS (e também regressões em componentes principais e a múltipla tradicional) na SG para pH da carne suína (medido aos 45min e às 24 horas após o abate), bem como identificar o número ótimo de componentes por meio dos métodos do grau de liberdade (GL) e validação cruzada (VC). O primeiro é baseado na Teoria de Informação Estatística e VC é empírica e fundamentada em amostras independentes do arquivo original. O PLS superou os demais métodos de regressão, fornecendo predições satisfatórias quando utilizadas apenas informações genotípicas (painel de SNP de baixa densidade). Além disso, os efeitos dos SNPs estimados via PLS possibilitaram identificar uma região relevante no cromossomo 4 que pode influenciar as características estudadas. Os métodos GL e VC foram similares quanto à determinação do número ótimo de componentes na análise PLS, porém o método GL pode ser recomendado devido a sua maior fundamentação estatística.

ABSTRACT: The main application of genomic selection (GS) is the early identification of genetically superior animals for traits difficult-to-measure or lately evaluated, such as meat pH (measured after slaughter). Because the number of markers in GS is generally larger than the number of genotyped animals and these markers are highly correlated owing to linkage disequilibrium, statistical methods based on dimensionality reduction have been proposed. Among them, the partial least squares (PLS) technique stands out, because of its simplicity and high predictive accuracy. However, choosing the optimal number of components remains a relevant issue for PLS applications. Thus, we applied PLS (and principal component and traditional multiple regression) techniques to GS for pork pH traits (with pH measured at 45min and 24h after slaughter) and also identified the optimal number of PLS components based on the degree-of-freedom (DoF) and cross-validation (CV) methods. The PLS method out performs the principal component and traditional multiple regression techniques, enabling satisfactory predictions for pork pH traits using only genotypic data (low-density SNP panel). Furthermore, the SNP marker estimates from PLS revealed a relevant region on chromosome 4, which may affect these traits. The DoF and CV methods showed similar results for determining the optimal number of components in PLS analysis; thus, from the statistical viewpoint, the DoF method should be preferred because of its theoretical background (based on the "statistical information theory"), whereas CV is an empirical method based on computational effort.

RESUMO Objetivou-se comparar os modelos multicaracterístico e de repetibilidade na estimação de parâmetros genéticos para as características número de leitões nascidos vivos (NLN) e às três semanas de idade (NL3), peso da leitegada ao nascimento (PLN) e às três semanas de idade (PL3), e peso médio do leitão ao nascimento (PMLN) e às três semanas de idade (PML3), considerando os três primeiros partos de fêmeas da raça Landrace. As estimativas de herdabilidade (h2) aumentaram até a terceira ordem de parto para as características PLN e PML3. Para NLN, NL3, PL3 e PMLN as h2 aumentaram da primeira para a segunda parição e reduziram da segunda para terceira parição. Em geral, as herdabilidades estimadas via modelo de repetibilidade foram menores que a média das estimativas obtidas utilizando o modelo muticaracterístico. As características PLN, PMLN e PML3 apresentaram altas correlações genéticas entre as diferentes parições (0,961–0,997), enquanto as características NLN, NL3 e PL3 (0,092–0,986) apresentaram valores irregulares de correlações genéticas entre as parições. Pelo critério de informação de Akaike corrigido o modelo de repetibilidade não foi indicado para a maioria das características estudadas. Esses resultados indicam que o modelo multicaracterístico é recomendado para avaliação genética das características número de leitões nascidos vivos e às três semanas de idade, peso da leitegada ao nascimento e às três semanas de idade e peso médio do leitão ao nascimento e às três semanas de idade, em diferentes parições, como características diferentes.

SUMMARY We aimed to compare multi-trait and repeatability models to estimate genetic parameters for the traits number of piglets born alive (NBA) and alive at 3 week of age (NP3), litter weight at birth (LW0) and at 3 week of age (LW3), and mean piglet weight at birth (MW0) and at 3 week of age (MW3), considering the first three farrowings of Landrace sows. Heritability (h2) estimates showed an increasing pattern up to the third farrowing for LW0 and MW3. For NBA, NP3, LW3, and MW0 h2 increased from the first to the second and decreased from the second to the third farrowing. In general, heritability estimated in the repeatability model was lower than the mean of the estimates in the multi-trait model. The traits LWO, MW0, and MW3 presented high genetic correlation among different farrowings (0.961–0.997), while NBA, NP3, and LW3 (0.092–0.986) presented irregular values among farrowings. The corrected Akaike information criterion shows that the repeatability model is not indicated for almost all of the studied traits. These results indicate that the multi-trait model is recommended for genetic evaluation of the traits number of piglets born alive and alive at 3 week of age, litter weight and mean piglet weight at birth and 3 week of age, in different farrowings, as different traits.

RESUMO: Objetivou-se estudar a efetividade da redução da dimensão da matriz de covariância do efeito genético direto na avaliação genética do crescimento (pesos dos 100 aos 730 dias de idade) de bovinos Simental, por meio da análise de componentes principais, e estimar componentes de (co)variância e parâmetros genéticos. A análise de componentes principais foi realizada ajsutando-se cinco diferentes modelos: um modelo multicaracterístico padrão, de posto completo, e quatro modelos de posto reduzido. Os modelos foram comparados via informação de Akaike (AIC) e informação Bayesiana de Schwarz (BIC). Os componentes de variância e parâmetros genéticos foram obtidos via REML. Os valores de AIC e BIC para os modelos testados foram similares, indicando a possibilidade da escolha de um modelo mais parcimonioso na avaliação genética da raça Simental. O primeiro componente principal explicou mais de 96% de toda variação genética aditiva direta em ambos os modelos. Os valores de herdabilidades foram maiores em idades mais avançadas e variaram de 0,05 (peso aos 100 dias) a 0,30 (peso aos 730 dias). As estimativas de correlações genéticas foram similares em todos os modelos e apresentaram mesma magnitude e comportamento independentemente do número de componentes principais adotado. Diante dos resultados, pode-se afirmar que apenas o primeiro componente principal foi suficiente para explicar quase que na totalidade a variação genética aditiva direta existente. Além disso, a similaridade dos parâmetros genéticos estimados e a menor demanda computacional são indicativos da possibilidade da utilização de modelos mais parcimoniosos na avaliação genética de bovinos Simental.

ABSTRACT: The aim of this research was to evaluate the dimensional reduction of additive direct genetic covariance matrices in genetic evaluations of growth traits (range 100-730 days) in Simmental cattle using principal components, as well as to estimate (co)variance components and genetic parameters. Principal component analyses were conducted for five different models-one full and four reduced-rank models. Models were compared using Akaike information (AIC) and Bayesian information (BIC) criteria. Variance components and genetic parameters were estimated by restricted maximum likelihood (REML). The AIC and BIC values were similar among models. This indicated that parsimonious models could be used in genetic evaluations in Simmental cattle. The first principal component explained more than 96% of total variance in both models. Heritability estimates were higher for advanced ages and varied from 0.05 (100 days) to 0.30 (730 days). Genetic correlation estimates were similar in both models regardless of magnitude and number of principal components. The first principal component was sufficient to explain almost all genetic variance. Furthermore, genetic parameter similarities and lower computational requirements allowed for parsimonious models in genetic evaluations of growth traits in Simmental cattle.

The objective of this study was to estimate genetic parameters and to evaluate the effects of inbreeding on productive and reproductive traits of dairy Gyr cattle. Single-trait animal models were used to estimate genetic parameters and solutions for inbreeding coefficients for milk (milk 305-d), fat (fat 305-d), protein (protein 305-d), lactose (lactose 305-d), and total solids (TS 305-d) yield up to 305 days of lactation, days in milk (DIM), age at first calving (AFC) and calving intervals (CI). The mean inbreeding coefficient was 2.82%. The models with linear and quadratic effects of inbreeding coefficients fitted the data better than the models without or with only linear effect of inbreeding coefficient for all traits. The increase in inbreeding coefficient caused several losses in productive and reproductive traits of dairy Gyr cattle. Estimates of heritability for milk 305-d, fat 305-d, protein 305-d, lactose 305-d, TS 305-d, DIM, AFC, and CI were 0.28, 0.27, 0.22, 0.21, 0.22, 0.17, 0.20, and 0.10, respectively. It is possible to achieve genetic progress in productive traits (especially in milk 305-d and fat 305-d) and age at first calving in dairy Gyr cattle through selection.

O objetivo neste estudo foi avaliar a estrutura genética da população de bovinos da raça Girolando no Brasil. Analisou-se o arquivo de pedigree de 26.969 animais, composto de 3.031 machos e 23.938 fêmeas. O nível de conteúdo de informação do pedigree na geração atual foi 61%, mostrando ser de qualidade moderada. O coeficiente de endogamia médio e o coeficiente de relação médio da população Girolando foram 0,11 e 0,13%, respectivamente. O tamanho efetivo da população, considerando a geração completa traçada, foi 188, acima do nível crítico. Do total de 9.457 ancestrais que contribuíram para a população de referência, 457 explicaram 50% da variabilidade genética da população. O número efetivo de fundadores foi 551 e o de ancestrais 393. O intervalo médio de geração foi de 5,26 anos, sendo ligeiramente maior nas trilhas gaméticas mãe-filho e pai-filha. A partir dos coeficientes estimados, pode-se concluir que a endogamia nos rebanhos da raça Girolando foi de pequena magnitude e que as práticas de acasalamento foram adequadas durante o período avaliado. No entanto, é importante continuar com o monitoramento desses coeficientes a fim de prevenir perda de variabilidade genética

The aim of this study was to evaluate the population structure of Girolando cattle in Brazil. The pedigree file contained 26,969 individuals, from which 3,031 were males and 23,938 were females. The average level of completeness of the pedigree in the current generation was of reasonable quality (61%). Inbreeding and average relatedness coefficients were low: 0.11 and 0.13%, respectively. Estimates of effective population size considering the full generations traced was 188, which is above the critical level range. The number of ancestors that contributed to the reference population was 9,457 animals, from which 457 explained 50% of the genetic variability of the population. The effective number of founders and the effective number of ancestors in this population were, respectively, 551 and 393. The average generation interval was 5.26 years, slightly higher in genetic pathways dam-son and sire-daughter. The inbreeding in the Girolando breed was of small magnitude, indicating that the current practices of mating were adequate during the study period. However, it is important to continue monitoring these coefficients in order to prevent loss of genetic variability

O objetivo deste trabalho foi apresentar modelagens alternativas, uni e bivariadas, para avaliação da conversão alimentar (CA) de suínos da raça Piau, com uso de inferência bayesiana. Os efeitos de sexo e genótipo sobre a CA dos animais foram avaliados por meio de procedimentos de simulação de Monte Carlo via cadeias de Markov (MCMC) e de integração aproximada aninhada de Laplace (INLA). O modelo univariado foi avaliado com diferentes distribuições para o erro - normal (gaussiana), t de Student, gama, log-normal e skew-normal -, enquanto, para o modelo bivariado, considerou-se o erro normal. A distribuição skew-normal foi o modelo mais parcimonioso para inferir sobre a resposta direta (univariada) da CA aos efeitos de sexo e genótipo, os quais não foram significativos. O modelo bivariado foi capaz de identificar diferenças significativas no ganho de peso e no consumo de ração em níveis de significância não detectados pelo modelo univariado. Além disso, ele também foi capaz de detectar diferenças entre sexos, quando agrupados por genótipos NN (machos, 2,73±0,04; fêmeas, 2,68±0,04) e Nn (machos, 2,70±0,07; fêmeas, 2,64±0,07), e revelou maior acurácia e precisão nas inferências nutricionais. Em ambas as abordagens, o método bayesiano mostra-se flexível e eficiente para a avaliação do desempenho nutricional dos animais.

The objective of this work was to present alternative uni- and bivariate modeling procedures for the evaluation of feed conversion (FC) of the Piau swine breed, using Bayesian inference. The effects of sex and genotype on animal FC were evaluated by the Markov chain Monte Carlo (MCMC) and the integrated nested Laplace approximation (INLA) procedures. The univariate model was evaluated using different distributions for the error - normal (Gaussian), t-Student, gamma, log-normal, and skew-normal -, whereas, for the bivariate model, the normal error was considered. The skew-normal distribution was the most parsimonious model to infer on the direct response (univariate) of FC to the effects of sex and genotype, which were nonsignificant. The bivariate model was capable to identify significant differences on weight gain and feed intake in significance levels not detected by the univariate model. Moreover, it was also able to detect differences between sexes, when grouped by NN (male, 2.73±0.04; female, 2.68±0.04) and Nn (male, 2.70±0.07; female, 2.64±0.07) genotypes, and revealed greater accuracy and precision for nutritional inferences. In both approaches, the Bayesian method proves flexible and efficient for assessing animal nutritional performance.